Handbook of Social Cognition: Volume 2: Applications [2, 2 ed.] 9780805810585

This edition of the Handbook follows the first edition by 10 years. The earlier edition was a promissory note, presaging

126 60 13MB

English Pages 514 [531] Year 1994

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Half Title
Title Page
Copyright Page
Table of Contents
Foreword
Preface
1. Stereotypes
2. Cognitive Processes in Attitude Change
3. Cognitive Perspective in Political Psychology
4. Some Cognitive Structures and Processes Relevant to Relationship Functioning
5. Social Cognition and Health Psychology
6. Social Cognition and Clinical Psychology: Anxiety, Depression, and the Processing of Social Information
7. On the Synergy Between Theory and Application: Social Cognition and Performance Appraisal
8. Consumer Judgment and Decision Processes
Author Index
Subject Index
Recommend Papers

Handbook of Social Cognition: Volume 2: Applications [2, 2 ed.]
 9780805810585

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

HANDBOOK OF SOCIAL COGNITION Second Edition

Volume 2 Applications

This page intentionally left blank

HANDBOOK OF SOCIAL COGNITION Second Edition

Volume 2 Applications Edited by

R obert S. W y e r , J r . T h o m as K . S rull University o f Illinois, Urbana-Champaign

ITT Psychology Press X

Taylor &. Francis Group NEW YORK A N D L O N D O N

First published 1994 by Lawrence Erlbaum Associates, Inc. Published 2014 by Psychology Press 711 Third Avenue, New York, NY 10017 and by Psychology Press 27 Church Road, Hove, East Sussex, BN3 2FA Psychology Press is an imprint o f the Taylor & Francis Group, an informa business Copyright © 1994, by Lawrence Erlbaum Associates, Inc. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data Handbook of social cognition / edited by Robert S. Wyer, Thomas K. Srull. — 2nd ed. p. cm. Includes bibliographical references and indexes. Contents: v. 1. Basic processes — v. 2. Applications. 1. Social perception. 2. Cognition. 3. Cognition—Handbooks. 4. Social perception—Handbooks. I. Wyer, Robert S., Jr. II. Srull, Thomas K. HM132.H333 1994 302 '.12—dc20 93-29484 CIP ISBN 13: 978-0-805-81058-5 (hbk)

Publisher’s Note The publisher has gone to great lengths to ensure the quality o f this reprint but points out that some imperfections in the original may be apparent.

Contents

Foreword Thomas M. Ostrom

vii

Preface

xiii

1.

2.

3.

4.

5.

6.

7.

Stereotypes David L. Hamilton and Jeffrey W. Sherman

1

Cognitive Processes in Attitude Change Richard E. Petty, Joseph R. Priester, and Duane T. Wegener

69

Cognitive Perspective in Political Psychology Shanto Iyengar and Victor Ottati

143

Some Cognitive Structures and Processes Relevant to Relationship Functioning Margaret S. Clark, Vicki S. Helgeson, Kristin Mickelson, and Sherri P. Pataki

189

Social Cognition and Health Psychology Leslie F. Clark

239

Social Cognition and Clinical Psychology: Anxiety, Depression, and the Processing of Social Information Gifford Weary and John A . Edwards

289

On the Synergy Between Theory and Application: Social Cognition and Performance Appraisal Jack M. Feldman

339 v

Vi

8.

CONTENTS

Consumer Judgment and Decision Processes Frank R. Kardes

399

Author Index

467

Subject Index

505

Foreword

The first edition of the Handbook o f Social Cognition was a revolutionary manifesto. It was both a challenge to traditional ways of social psychological the­ orizing and a statement from the rabble that “we are here and we mean to stay.” It was confrontational and passionate. It was fists and sinew demanding recogni­ tion and acceptance.

THE GROWTH OF SOCIAL COGNITION

That first edition was published only a decade ago, in 1984. This is an unusually short interval of time between editions for volumes of this sort. For example, the Handbook of Social Psychology has averaged over 15 years between editions, and yet, there clearly was a need for this revised edition. The need was motivat­ ed not by discontent with the coverage or scholarship of the first edition, but rather by the enormous outpouring of new research and theory on the topic in the past 10 years. The field of social cognition has changed substantially since the ap­ pearance of the first edition. The rapid expansion of social cognition also can be seen in the large number of edited volumes that have been published in the past 10 years, including the Advances in Social Cognition series, edited by Wyer and Srull, which focus on specialized topics in the field. In addition, the first comprehensive text on social cognition, authored by Fiske and Taylor, was published in the same year as the first edition of this Handbook. The second edition of that text, involving re­ vii

viii

FOREWORD

visions ranging from 40% to 80% of each chapter, was published in 1991. Those authors, like the editors of this Handbook, recognized and responded promptly to the fast-paced dynamism of the field. The table of contents of the present edition of the Handbook reveals several clues as to the nature of the changes in this field. Of the 15 chapters in this sec­ ond edition, only four involve authors who had contributed to the first edition. This illustrates several points. First, it displays the rapid growth in the number of scholars entering the field of social cognition. Second, it shows the increased acceptance of social cognition concepts as relevant to understanding traditionally important social psychological processes (see Volume 1) and domain-specific ap­ plications of those processes (see Volume 2). Further comparison of contributors shows that, although 7 of the 17 chapters in the first edition were authored or coauthored by cognitive psychologists, no authors in the present edition are identified primarily with cognitive psychology. In assembling the first edition, the editors apparently believed it necessary to recruit cognitive psychologists for the purpose of educating social psychologists about many of the basic principles and issues in cognitive psychology. In contrast, the present edition reflects a confidence that most readers come to the volume with such understandings well in hand. It correctly takes for granted that the current generation of readers will have already completed their basic training in cogni­ tive psychology through undergraduate and graduate course work. The present edition displays a comfortable self-acceptance with the role that social cognition has come to play in the field.

THE SOVEREIGNTY OF SOCIAL COGNITION

The first edition of this Handbook led off with a chapter entitled “The Sovereignty of Social Cognition.” No other chapter conveyed the revolutionary manifesto of that edition more forcefully than did this one. The chapter, and especially its zealous title, proved offensive to some scholars in the field. It was viewed as being, in part, naive and, in part, delusional. Many of those skeptics felt that social cognition would prove to be only a minor (and slightly offensive) digres­ sion in the history of social psychology. Perhaps the title of that chapter was ill chosen. It certainly led some of the chapter’s critics to not even bother reading the chapter. Rather, those schol­ ars condemned the whole field on the basis of the presumptuousness of the chapter’s title. Yet the substance of the chapter was intended to be self-critical and conciliatory, rather than arrogant and confrontational. It was intended to embrace diverse traditional perspectives rather than to reject or condemn them. The editors of this second edition suggested to me that I prepare a new up­ dated introductory chapter. After some thought, I declined. It seemed to me

FOREWORD

ix

that the basic points made in the earlier chapter were still valid and that, although the exposition of those ideas could be improved, the issues raised there are still the ones most fundamental to the field of social cognition. To obtain a sense of the scope and integrative potential of social cognition, contemporary readers may find it beneficial to include that chapter as part of their readings in the field. Three topics received special emphasis in that chapter. I offer some commen­ tary on each in the remainder of this Foreword.

DEFINING SOCIAL COGNITION

When that chapter was written, there had been 22 separate attempts to define the term social cognition. That number probably has increased to over 100 by now, given the number of papers, chapters, and textbooks that have been pub­ lished in the interim on the subject. An analytic synthesis of this extensive list would prove overwhelming and probably would yield little of value even if the required time and effort were devoted to the task. After all, the real goal of the field is to improve one’s understanding of social psychological phenomena. There is no reason to be distracted by definitional squabbles that dwell only on the label given to the new research and theory. I regard single-sentence definitions of social cognition to be slightly offen­ sive, despite their necessity for textbooks and despite the fact that I have done so in my own writings. It is impossible to convey past history and future aspira­ tions in a single sentence, it is impossible to capture the creative richness of the ideas in a single sentence, and it is impossible to portray the enthusiasms of its scholars in a single sentence. I stated in the earlier chapter that my definition of the field was the entire chapter. It would be more accurate to say that my preferred definition is the entire Handbook o f Social Cognition, both this and the first edition combined. At the heart of social cognition is the conceptual orientation that has emerged from the information-processing perspective in cognitive psychology, a per­ spective that recently has expanded to include cognitive science. The social cog­ nition approach is based on the conviction that constructs relevant to cognitive representation and process are fundamental to understanding all human responses, regardless of whether those responses are social or nonsocial in nature. Cog­ nitive psychologists have applied these concepts to the analysis of a wide range of phenomena, such as text comprehension, recall, recognition, classification, reasoning, vision, and audition. Social cognition researchers share this theoretical perspective, differing solely in the phenomena to be understood. Those social phenomena are, of course, the focus of the two volumes in this Hand­ book.

X

FOREWORD

ACTION IS THE CORE OF SOCIAL BEHAVIOR

The single most important theme of my earlier chapter was that cognitive activi­ ty was in the service of interpersonal action. Thinking, along with its more mind­ less cognitive companions, has evolved to allow persons to act on and react to a world that is first and foremost social in nature. Lamentably, it seems that this theme was poorly understood and widely ignored by the field. Despite this neglect, I believe the analyses offered in that chapter remain valid today. The chapter presented a fourfold analysis of the difference between social and nonsocial knowledge and their attendant cognitive processes. Much of the research covered in the first two editions of the Handbook falls in the first three categories of that classification system. Unfortunately, these three categories are the least social of the system. They describe characteristics of persons as stimulus objects (Category One), characteristics of the perceiver (Category Two), and contingen­ cies in human interaction (Category Three). The nature of these categories, and their relation to earlier attempts at defining social cognition, were discussed at some length in that chapter. Category Four highlights the perceiver as an interaction participant. It includes the processes of the first three categories, but goes beyond them. In doing so, it aspires to understand cognitive representations and cognitive processes as they operate on-line while the person engages in meaningful congress with one or more other persons. Category Four information processing has several important features. Most notably, it involves a heavy processing load on the perceiver’s cognitive capaci­ ties. As perceivers, we have personal goals to form and modify on-line. We have to be selectively attentive to the other person’s responses, and to draw inferences from those responses regarding the other person’s beliefs, intentions, and affec­ tive dispositions. We must be able to access memory for past encounters with that person as well as earlier transactions in the current encounter. We must be able to anticipate the partner’s reaction to alternative personal initiatives and to select for display the one judged to be most effective. Not only do social exchanges have multiple sources of demand on our cogni­ tive system, but they also possess a number of very distracting features. Distrac­ tions come from the inherent unpredictability of our partner’s responses, from the fact that concurrent verbal and nonverbal behavior of the partner may be in­ consistent with one another, and, in the case of multiple interaction partners, from the unpredictability of who will be responding next. Category Four processing appears oppressive. It is seemingly overwhelming in the number and nature of demands on the cognitive system. Yet humans have evolved to deal with such demands on most occasions in a near effortless man­ ner. The real challenge to social cognition, then, is to understand how the many components of the information-processing system work together to support hu­ man action, especially social interaction.

FOREWORD

Xi

One important research vector has emerged that directly informs Category Four information processing. The heavy cognitive demands during face-toface interaction require that social behavior be guided by automatic rather than controlled processes—on on-line rather than retrieval-based processes. Deliber­ ative processes, ones that involve selective retrieval from long-term memory and thoughtful analyses of alternatives, are serial in nature, easily disrupted, and relatively slow in execution. Therefore, such deliberative processes can­ not account for most of what guides action in interaction contexts. More needs to be learned about how automatic processes operate in producing the multi­ ple, simultaneous, verbal, and nonverbal behaviors that occur during inter­ action. More needs to be known about the interface between automatic and controlled processes, especially in regard to controlled override of automatic inclinations.

IS SOCIAL COGNITION SOVEREIGN?

Of course it is. The earlier chapter presented six arguments in support of the sovereignty proposition. They dealt with issues such as the newly achieved relevance of cognitive research to traditional social phenomena, to the priority of social over nonsocial processes, to the capacity of cognitive principles to help understand affective and motivational processes, and to the potential for linking cognitive constructs to the neurological substrate. Not only do those arguments remain valid today, but they have been strengthened by work that has appeared subsequent to the chapter’s publication. The present edition adds a seventh argument supporting the ultimate destiny of social cognition as monarch of social psychology. In its early stages, social cognition was dismissed as the sterile study of memory under laboratory condi­ tions. This shallow caricature was accepted by many scholars in social psycholo­ gy who were concerned with applying principles of their field to pressing social, organizational, and economic problems. Acceptance of that view led these schol­ ars initially to dismiss social cognition on the grounds that it was irrelevant to their concerns. A great strength of this second edition is its inclusion of six new chapters that cover many of the ways in which social cognition has contributed to an improved understanding of, and solutions to, such concerns. These chapters cover clinical psychology, close relationships, political psychology, consumer psychology, per­ sonnel psychology, and health psychology. The pervasiveness of the cognitive orientation is such that it has permeated all areas of applied social psychology. For example, other chapters could have been included on topics such as law and psychology, the psychology of women, the psychology of ethnic minority issues, and a number of other fields that are now divisions of the American Psychologi­ cal Association.

Xii

FOREWORD

It is clear that social cognition is no longer revolutionary. Instead, it has at­ tained the status of “standard science.” It no longer can be taught to graduate stu­ dents as an exciting new conceptual orientation. Students arrive having had broad exposure to the basic principles in their undergraduate curriculum, and therefore are baffled when they encounter fervent advocacy and claims of novelty. This second edition of the Handbook fully conveys the new maturity of social cognition. It amply demonstrates the relevance of information-processing princi­ ples to understanding the cognitive dynamics underlying a broad array of tradi­ tional domains in social psychology. Indeed, it is easy to envision a future in which there is no longer a need for a separate Handbook o f Social Cognition. The Handbook o f Social Cognition will become the Handbook o f Social Psychology. —Thomas M. Ostrom Ohio State University

Preface

When the first edition of the Handbook of Social Cognition appeared in 1984, it was a promissory note. The field was then in its infancy, and the areas of research and theory that came to dominate the field during the next decade were only be­ ginning to emerge. The concepts and methods used had frequently been borrowed from cognitive psychology, and had been applied to phenomena in a very limited number of areas. Nevertheless, social cognition promised to develop rapidly into an important area of psychological inquiry that would ultimately have an impact not only on several areas of psychology but on other fields as well. The promises made by the earlier edition of the Handbook have generally been fulfilled. Since the publication of the first edition 10 years ago, social cognition has become one of the most active areas of research in the entire field of psychol­ ogy, whose influence has extended not only to many other subareas (health psy­ chology, clinical psychology, personality, etc.) but to totally different disciplines (political science, marketing and consumer behavior, organizational behavior, etc.). The impact of social cognition theory and research within a very short period of time is incontrovertible. The current edition of the Handbook clearly conveys the advances that have been made over the years. It is divided into two volumes. Volume 1 is devoted to research and theory pertaining to basic components of social information processing: the encoding of information, its representation in memory, infence processes, and response selection. Although several of these topics were covered in the first edition of the Handbook as well, the content of the present chapters is virtually nonoverlapping. For example, John Bargh, who also authored a chapter xiii

Xiv

PREFACE

in the earlier edition on automatic versus controlled information processing, has identified four separate components of automaticity that had not been distinguished at the time the earlier chapter was written, each of which is a distinct focus of current research and theory. Two of the following three chapters, by Robert Wyer and Donal Carlston on the mental representation of people and social events, and by John Kihlstrom and Stanley Klein on the cognitive representation of self, also deal with topics that were discussed in the earlier edition. But, like Bargh’s chap­ ter, they cover areas of research and theory that were largely unanticipated by the earlier articles. The fourth chapter, by Eliot Smith, discusses the role of procedural knowledge in social information processing—an exceptionally impor­ tant topic that had not even been clearly identified at the time the earlier Hand­ book was published. The next three chapters in Volume 1 also deal with topics that were covered in the first edition but reflect the major advances that have been made in the in­ terim. The chapter on social inference by Denise Beike and Steven Sherman replaces a narrower analysis of cognitive heuristics, incorporating them into a more general discussion of judgment phenomena. Fritz Strack’s chapter on response processes reviews research on the cognitive factors that underlie the interpretation of requests for information in a social context that has implications for responses in both laboratory and nonlaboratory situations. Gerald Clore, Norbert Schwarz, and Michael Conway provide a detailed analysis of both the cog­ nitive antecedents of affect and emotion and the impact of affective reactions, once elicited, on the processing of new information. Few, if any, of the issues discussed in either of these chapters had been identified a decade ago. Volume 2 of the Handbook reviews information processing approaches to un­ derstanding phenomena in a number of other specific domains of inquiry, none of which was considered in the earlier edition. The research in these domains has often been stimulated by theory and concepts discussed in Volume 1. However, this research has often led to the identification of new theoretical and empirical issues whose implications extend beyond the specific areas in which they were identified. The first chapter, for example, by David Hamilton and Jeffrey Sher­ man, concerns a phenomenon to which several of the processes discussed in the first volume are relevant. The development and use of stereotypes is neverthe­ less a fundamental topic in social cognition that was not explicitly considered in the earlier edition at all. Chapter 2, by Richard Petty, Joseph Priester, and Duane Wegener, provides an exceptionally broad and sophisticated review of in­ formation processing approaches to understanding communication and persua­ sion. Shanto Iyengar and Victor Ottati then give an equally penetrating analysis of the cognitive and affective bases of political judgment. This is followed by reviews of cognitive approaches to (a) the dynamics of close relationships, by Margaret Clark and her colleagues, (b) health psychology, by Leslie Clark, and (c) clinical psychology (with a special emphasis on depression) by Gifford Weary and John Edwards. The final two chapters, by Jack Feldman and Frank Kardes,

PREFACE

XV

review and evaluate the role of information processing in personnel appraisal and consumer behavior, respectively. The chapters in this volume combine to demon­ strate the important influence that social cognition has had not only on other areas of psychology but on other disciplines as well. There are inevitably limitations to the present volume. As Thomas Ostrom points out in the Foreword, an ultimate objective of social cognition research is to understand the cognitive processing that occurs in social interaction situa­ tions of the sort in which information is often exchanged outside the laboratory. At this writing, social cognition research has only begun to examine in detail many of the social contextual factors that underlie the interpretation of informa­ tion and responses to it more generally. To this extent, the present Handbook, like its predecessor, is a promissory note, providing the theoretical and empiri­ cal foundations of research to be performed in the future. These future research directions are indicated in virtually all chapters of the Handbook in the context of evaluating our current state of knowledge. The success of a reference volume depends on both the quality of the contri­ butions contained in it and their timeliness. The contributors to the Handbook are all among the foremost scholars in the field in the areas to which their chap­ ters pertain. We are greatly indebted to them, not only for the quality of their contributions but also for their cooperation in preparing their chapters under se­ vere time constraints despite numerous other demands and obligations. Many others have also contributed to the success of this project. Most obviously, the editors and staff of Lawrence Erlbaum Associates, and Larry Erlbaum himself, have been constant sources of support since the beginning of the project, and have been instrumental in facilitating its coming to fruition. Our association with LEA on this project, as in the past, has been immensely gratifying, and we can­ not imagine a better publisher with whom to work. Robert S. Wyer, Jr. Thomas K. Srull

This page intentionally left blank

1

Stereotypes

David L. Hamilton Jeffrey W. Sherman University o f California, Santa Barbara

Contents Cognitive Processes in Stereotype Formation 3 Categorization 4 Illusory Correlation 11 Stereotypes as Cognitive Structures 15 Early Approaches 15 Cognitive Representations of Groups and Group Members Representation of Group Variability Information 25 Stereotyping and Information Processing 33 Stereotypes: Encoding and Retrieval Biases 33 Stereotyping and Inference 37 The Automaticity of Stereotyping 40 Affect, Cognition, and Stereotyping 42 Incidental Affect and Stereotyping 43 Integral Affect and Stereotyping 45 Stereotype Change 47 Models of Stereotype Change 49 Research Evidence 50 A Final Comment 57 Acknowledgments 57 References 58

15

What is a stereotype? How do people develop stereotypes, and why? Why are stereotypes so prevalent and persistent? When and how are stereotypes used? How can we change people’s stereotypes?

These are questions that all of us—lay people and social scientists alike—have pondered when we think about stereotypes and their effects on our perceptions

1

2

HAMILTON AND SHERMAN

of various ethnic groups. Although seemingly simple, these questions cut to the heart of the nature and functioning of stereotypes. They are questions of both societal and scientific concern and have been for a long time. Despite their im­ portance, neither society nor science has dealt adequately with the issues and problems that these questions raise. This chapter assesses what we have learned about some of those issues from social psychological research, and particularly from research guided by a social cognition approach to the topic. Our review and analysis is organized around questions much like those posed earlier. However, we shall see that, when ap­ proached scientifically, these broad questions actually mask a number of more specific questions, each of which presents its own issues for analysis. A decade ago, Ashmore and Del Boca (1981) provided a useful summary of three conceptual approaches that have guided past theorizing and research on stereotyping and intergroup perceptions. They identified these approaches as the psychodynamic, sociocultural, and cognitive orientations. The psychodynamic approach emphasizes the role of motivational forces and psychological benefits that can lead to and perpetuate the use of stereotypes. With its roots in Freudian thinking, this perspective includes the use of defense mecha­ nisms such as projection and displacement of self-related sources of tension onto others, scapegoating, and an emphasis on how early childhood experiences af­ fect intrapsychic needs in explaining intergroup perceptions. The sociocultural approach focuses on the variety of means by which intergroup beliefs and atti­ tudes are acquired and maintained through social learning and social reinforce­ ments. The focus is on how stereotypes and prejudice can be learned and perpetuated through socialization experiences, peer group influence, and media portrayals. The cognitive approach views stereotypes as belief systems or cogni­ tive structures that can guide information processing, and it examines how those structures arise and how their influence on information processing affects per­ ceptions of and interactions with members of stereotyped groups. These orientations offer complementary, rather than competing, explanations for various phenomena involving intergroup perception. Therefore, it is likely that any single orientation is limited to providing only a partial account of these phenomena (Hamilton & Trolier, 1986; Stroebe & Insko, 1989). Nevertheless, most research on how stereotypes form and function has been guided by the con­ cepts and methods of one of these approaches. In recent years, the cognitive ap­ proach has been particularly active and influential in generating such research. Like the other contributions to this handbook, this chapter emphasizes the ad­ vances in understanding that have been generated by an information processing analysis. However, as a means of placing this work in context, at several points we include commentary reflecting the other orientations and contrasting their em­ phases with those of the cognitive approach. From the cognitive perspective, a stereotype can be defined as “a cognitive structure that contains the perceiver’s knowledge, beliefs, and expectations about

1.

STEREOTYPES

3

a human group” (Hamilton & Trolier, 1986, p. 133). Stereotypes are abstract knowledge structures linking a social group to a set of traits or behavioral charac­ teristics. As such, stereotypes act as expectancies that guide the processing of information about the group as a whole and about particular group members (Hamilton, Sherman, & Ruvolo, 1990). In addition to these generalized expect­ ancies, one’s knowledge about particular group members (or exemplars) also may influence judgments about groups and their members.

COGNITIVE PROCESSES IN STEREOTYPE FORMATION

How do people develop stereotypes? We suspect that most people, if asked this question, would respond in one of two ways. On the one hand, they might point to the important influence of parents and other significant figures (e.g., teachers, peer groups) on the attitudes and beliefs that people develop in their formative years. On the other hand, they might cite the impact of public media, noting that people come to believe what they see about various ethnic groups as they are portrayed on television and in movies. These would not be unreasonable answers to the question. In fact, these mecha­ nisms are at the heart of sociocultural explanations of stereotyping. The seeds of people’s conceptions of various racial and gender groups are planted in early child­ hood by influential adults in their lives, and they are fostered and perpetuated through their repeated perceptions of members of these groups in certain social roles as they are portrayed in the media. The role of social learning processes in the formation and maintenance of stereotypes has been a major focus of inter­ group research for a long time. In addition to these processes, the cognitive approach to stereotyping has fo­ cused on other mechanisms that can contribute to the initial formation of stereo­ typic belief systems. Although this approach also has a long history, it has been the catalyst for an enormous amount of research during the last 15 years. This resurgence is due, at least in part, to several empirical discoveries during the 1970s that pointed to cognitive mechanisms and biases that, in and of themselves, could contribute to the formation of stereotypes, independently of actual inter­ group conflict or of social influences from significant others. Stereotyping depends on the perception that a group of persons comprises a meaningful social entity. If individual persons were not perceived as belonging to some social unit, then there would be no basis for developing a stereotype. Inherent in this process is the perceptual separation of different social categories. Hence, stereotyping begins with the perceptual differentiation between groups of persons. That differentiation does not, in itself, mean that a stereotype will be formed; we know, for example, that there are blue-eyed and brown-eyed peo­ ple, but we do not have rich stereotypes associated with these groups. The for­ mation of the stereotype involves an additional step of associating certain attributes

4

HAMILTON AND SHERMAN

or features with those differentially perceived groups. That is, we develop be­ liefs about the attributes that are characteristic of each group and therefore dis­ tinguish between the groups. But these belief systems will not form if we have not already identified a collection of individuals as a group and differentiated those persons from some others. Therefore, any process that contributes to the differen­ tiation between groups constitutes a potential basis for the formation of stereo­ types. One of the key contributing factors to the resurgence of the cognitive approach to stereotypes was the fact that several lines of research provided evi­ dence of how cognitive mechanisms can contribute to this process. Categorization

In their perceptions of others, people often “see” others not (or not solely) as in­ dividual persons, but rather as members of social groups. Given that each person belongs to numerous social groups (based on gender, race, nationality, religion, occupation and socioeconomic status, political orientation, lifestyle, interests, etc.), viewing others in terms of such category memberships certainly captures some of the important elements of social structure and social life. In that sense, categoriz­ ing others into groups simply reflects social reality. However, beyond that, research has shown that aspects of people’s cognitive mechanisms and function­ ing contribute to and derive from this categorization process (Hamilton & Trolier, 1986; Miller & Brewer, 1986; Oakes & Turner, 1990; Taylor, 1981). These mechanisms can have important implications for people’s perceptions of and be­ havior toward group members. The important role of categorization in stereotyping has been recognized for many decades. It was implicit in Lippmann’s (1922) insightful analysis of the use of stereotypes in perceptions of groups. Its role was emphasized explicitly by Allport (1954) in his classic analysis of stereotyping and prejudice. But it was the simultaneous and independent development of two quite different lines of work in the early 1970s that led to an explosion of research on the role of categoriza­ tion in intergroup perception. One was the development of cognitive psychology and its focus on how acquired information is organized and stored in terms of long-term “knowledge structures.” The other was a program of research initiated by Tajfel (1969, 1970) that provided empirical documentation of the fundamen­ tal impact of the categorization process on social perception and behavior. Both of these initiatives focused attention on the role of cognitive factors in stereotyp­ ing and intergroup perception. To understand the role of categorization in stereotyping, several questions need to be addressed. First, why do people categorize others into groups at all? Why not simply perceive and understand them as individuals? Second, what are the bases of social categorization? What determines the particular categorization(s) that will be used in any given situation? Third, when will the perceiver catego­ rize others according to social groups, and when will others be perceived as

1.

STEREOTYPES

5

individual persons? The pervasive importance of the categorization process be­ comes apparent as these and other issues are discussed in this chapter. We begin with the question of why people categorize others into groups. Why Categorize?

Why would perceivers overlook the individuality of the persons they encoun­ ter and move instead to viewing them in terms of social categories? Two major forces driving such categorization are prevalent in the literature. One emphasis views categorization as a cognitive mechanism serving the informational needs of the perceiver. Specifically, the perceiver (a) must use a limited cognitive processing system to cope with a rich and complex social stimulus environment, yet (b) needs to understand and anticipate interactions with that environment. Categorization can facilitate meeting both of those demands. The second empha­ sis views social categorization as deriving from people’s desire to evaluate them­ selves positively, and therefore as motivated to see their own group as different from—and better than—other social groups. We consider each of these bases of categorization in the following subsections. Categorization as Cognitive Efficiency. One reason to categorize others based on their apparent similarities derives from the sheer complexity of the social en­ vironment. The richness of social stimulation provided by that environment places processing demands on the human cognitive system. Therefore, attention is direct­ ed at some aspects of the social environment while others are ignored. Simplify­ ing strategies for dealing with this information overload become functionally adaptive. As a consequence, perceivers group objects in their stimulus world into categories on the basis of their similarities and differences. Thus, categorization can be a response to the demands of information overload. This function of perceiving others in terms of groups was demonstrated by Rothbart, Fulero, Jensen, Howard, and Birrell (1978). Subjects were presented information describing the attributes of individuals and were subsequently asked to rate the group composed of these persons on a series of attributes. The stimu­ lus items (person-attribute pairings) were arranged such that the experimenters could determine whether subjects’ ratings of the group were based on an accumu­ lation of their conceptions of individual persons or, alternatively, on their con­ ception of the group as a whole. When the number of stimulus items was relatively small (low memory load condition), subjects’ judgments indicated that they or­ ganized the stimulus information in terms of the individual persons described. In contrast, when a large number of stimulus descriptions was presented (high memory load condition), subjects organized the descriptive information in terms of the group as a whole. Thus, under conditions of strained capacity, perceivers were less likely to develop person-based conceptions than to establish a categori­ cal representation at the group level.

6

HAMILTON AND SHERMAN

According to this cognitive view, then, categorization is, in part, a response to information overload, serving to simplify the perceiver’s processing task. In this sense, categorization involves information loss. But this is not the whole sto­ ry. An important consequence of categorization is that it also affords informa­ tion gain. That is, through categorization, persons typically are perceived in terms of social groups about whom the perceiver, through past experiences and social learning, has developed knowledge and beliefs. This accumulated knowledge and beliefs can then be applied in understanding individual group members through inference processes. Assuming that these representations are veridical to some extent (or at least might be “functionally accurate”; Swann, 1984), these stereotypebased elaborations are likely to be useful in going beyond the information avail­ able. Thus, from the cognitive perspective, categorization involves both infor­ mation loss and information gain. Categorization as Self-Enhancement. In addition to these cognitive mecha­ nisms that promote categorization, several motivational factors have been proposed that contribute to this same process (Maass & Schaller, 1991; Stangor & Ford, 1992; Stroebe & Insko, 1989). The most prominent theoretical account emphasiz­ ing motivational roots of categorization is social identity theory (Tajfel & Tur­ ner, 1979, 1986; see also Turner, 1987). The central hypothesis of this theory is that a person’s self-esteem is, in part, derived from his or her membership in social groups. Because people typically want to maintain positive self-regard, they are motivated to hold favorable evaluations of the groups to which they belong. But there is no objective yardstick for gauging the desirability of any particular social group; such evaluations are inherently subjective. Therefore, people en­ hance their own group’s favorability by psychologically establishing its relative superiority in comparison with some out-group. Thus, people are motivated to accentuate the evaluative difference between in-group and out-group, thereby creat­ ing intergroup discrimination based on the desire to maintain positive identity. The development of social identity theory was stimulated by the consistent finding of in-group bias—that people evaluate their in-groups more favorably than out-groups, even when the intergroup distinction is arbitrary or based on a trivial criterion (Tajfel, 1970; see later section on in-group/out-group differentiation). Evidence for this effect is pervasive (Brewer, 1979; Messick & Mackie, 1989) and consistent with social identity theory, although the important role of self­ esteem maintenance in producing these intergroup effects has not been established clearly by research findings (Maass & Schaller, 1991; Messick & Mackie, 1989). Bases of Social Categorization

As you are on your way to work early one morning, you notice a group of joggers running through a park. One of them is an attractive, 30-ish woman, well attired in her sweatshirt and sweatpants, headband, and running shoes. In noticing

1.

STEREOTYPES

7

her, you might think of this woman in terms of any of several social categories to which she might belong. For example, you might categorize her as a woman, or as a runner, or more specifically, as a female runner. From her age group and expensive sportswear, you might regard her as a “yuppie.” Or, in more general terms, you might think of her as “a person who tries to stay physically fit.” Any of these might be plausible—and reasonably accurate—categorizations. On the other hand, you are unlikely to immediately think of her as “a resident of my community” (which also would be accurate). This example illustrates that each person is a member of numerous social group­ ings, any one of which might serve as an appropriate categorical basis for per­ ceiving the person. What determines which of several possible categorizations will be used in one’s perception of this person? At this point, it is impossible to provide a definitive answer to this question, but several possibilities illustrate the range of alternative bases for categorization at the perceiver’s disposal. Primitive Categories. Some theorists (Bower & Karlin, 1974; Brewer, 1988; Bruner, 1957; Fiske & Neuberg, 1990) have suggested that a small number of critical or primitive categories are used automatically and universally in perceiv­ ing other human beings. For example, when we encounter a person it seems nearly impossible not to notice the person’s gender, race, and age group, these being the three most commonly cited candidates for primitive categories. Why would certain categories achieve this status of “primitive” categories that are employed automatically and universally? There are several highly interrelat­ ed reasons, each of which has some plausibility but has difficulties as well. First, the widespread use of these categories might simply reflect the fact that these are broad categories of human beings, which can serve as a basis for finer dis­ tinctions within these primitive categories. However, this view has difficulty ex­ plaining why certain universal features (e.g., gender, race) become the basis for primitive categorization, whereas others (e.g., eye color, hair color) do not. Also, some research suggests that once different levels of categorization or subtypes have been established, perceivers will not rely on the broader higher level clas­ sifications in categorizing others, but instead will use more specific “basic level” categories (Brewer, Dull, & Lui, 1981; Devine & Baker, 1991). Second, primitive categories may have achieved their great importance be­ cause they have primacy status. That is, they are the first categorizations per­ ceivers can make because they reflect features that usually are salient in a person’s appearance and hence are immediately obvious to the perceiver. This explana­ tion gives special importance to early experiences and to the first features people see in others, but neglects the importance of more recent experiences and recen­ cy effects. Finally, perhaps the key element in this view is the argument that primitive categories are in fact the categories of greatest importance for capturing signifi­ cant information about others. The idea here is that there are certain human charac­

8

HAMILTON AND SHERMAN

teristics that reflect fundamental distinctions in social behavior, and it is there­ fore functionally useful for the perceiver to categorize others in terms of those distinctions. For example, if there are meaningful differences in social interac­ tions with males versus females, or older versus younger persons, then recogniz­ ing another person’s status on these dimensions would be useful in anticipating the nature of those interactions. The difficulty in substantiating this argument, of course, lies in documenting that these features are, in fact, most predictive of differences in social behavior. Whether these categories have any special status as “primitive” categories or not, they will be used frequently as information about others is encoded. Fre­ quency of category use is one of the primary determinants of that category’s use in the future; use of a category increases its accessibility for future categoriza­ tion (Bargh, Bond, Lombardi, & Tota, 1986; Higgins, Bargh, & Lombardi, 1985; Srull & Wyer, 1979, 1980). With repeated use across time and context, such cognitive procedures can become so routine (or “proceduralized”) that they are virtually automatic (Smith, 1990). In-group/Out-group Differentiation. Another fundamental basis for social categorization is the distinction between groups to which one belongs (in-groups) and those to which one does not (out-groups). The importance of in-group/out­ group differentiations for understanding intergroup conflict and behavior has been recognized in social psychology for a long time (Allport, 1954; Campbell, 1965; Sherif, 1967). Although some writers (Allport, 1954) emphasized categorization as an important element in intergroup perception, it was only with the work of Tajfel (1969, 1970; Billig & Tajfel, 1973; Tajfel, Billig, Bundy, & Flament, 1971) that the central role of the categorization process was documented dramatically in empirical research. Tajfel established a “minimal intergroup paradigm” (see Diehl, 1990) for study­ ing the influence of social categorization processes, independent of actual inter­ group conflict. In this paradigm, the experimental procedure occurs in two phases. In the first phase, several subjects make individual judgments about a number of stimuli (e.g., evaluative ratings of paintings by two artists; estimations of the number of dots in a stimulus display) and subsequently are given feedback that identifies them as similar to some of the other subjects and as different from the rest (e.g., in their preferences for one of the painters, or in tending to overesti­ mate the number of dots). In actuality, subjects are randomly assigned to one or the other group, and this is done in such a way that subjects do not know which other subjects are members of their own or the other group. In this way, two groups of subjects are formed on the basis of a task that (a) seemingly might represent some “real” psychological difference yet (b) seemingly would be un­ related to performance on the tasks in the second phase of the study. In this sec­ ond phase, subjects are asked to make evaluative ratings of members of their own and the other group or are asked to allocate resources to members of their

1.

STEREOTYPES

9

own and the other group. The primary finding from numerous studies using this paradigm is in-group bias, such that subjects allocate more resources and evalu­ ate more favorably members of their own group, even though they do not know the specific identity (only the group membership) of the persons about whom they are responding. Initial findings of the Tajfel group became the catalyst for subsequent research that has continued for 20 years on a variety of related topics, including inter­ group differentiation, in-group bias, the perception of variability within groups, and the out-group homogeneity effect (for recent reviews, see Diehl, 1990; Mes­ sick & Mackie, 1989; Vanbeselaere, 1991). These findings also have served as the basis for several important theoretical developments in the intergroup litera­ ture (Miller & Brewer, 1986; Tajfel & Turner, 1986; Turner, 1987). Context-Based Differentiation. One of social psychology’s primary themes is the importance of the social context on people’s thinking, perceptions, and be­ havior. It is not surprising, then, that research has shown that the nature of the social context can influence the way group members are perceived. Specifically, categorization can be based on whatever features happen to be salient within the stimulus context. This effect was demonstrated with nonsocial stimuli by Tajfel and Wilkes (1963), who showed subjects a series of eight lines of increasing lengths. In one condition, the four shorter lines were labeled A, whereas the four longer lines were labeled B. In another condition, no labels were associated with the lines. The subjects’ task was to estimate the length of each line. Relative to the no-label condition, subjects in the label condition overestimated the difference in length between the lines of the two categories (i.e., they exaggerated the difference be­ tween the fourth and fifth lines). Similar effects have been demonstrated in the social domain. For example, Wilder (1978) showed that perceivers assume greater belief similarity between two members of the same group than of different groups, even though the basis of group membership is not diagnostic for the belief judgment being made (see also Wilder, 1981). Research by Taylor (1981; Taylor, Fiske, Etcoff, & Ruderman, 1978) showed that salient features (race, gender) can become the basis for processing and storing information about group members. For example, in one study, subjects observed a discussion among three Black and three White males and later were asked to identify which person had made certain comments during the discussion. Subjects made more within-category than between-category er­ rors on this matching task. These findings suggest that salient race cues induced social categorization, which in turn influenced subjects’ processing and represen­ tation in memory of the information about the participants. Stimulus Salience. The immediate social context also can make a person dis­ tinctive. A person of one race or gender, for example, may be in a group other­ wise composed of persons belonging to a different race or gender. Alternatively,

10

HAMILTON AND SHERMAN

the social context may exist in the perceiver’s head, in the form of a priori or normative expectations, such that a person becomes salient by deviating from one’s stereotypic conception of that person’s group. In any event, such salience, however acquired, has consequences for how that salient person is perceived. Taylor and her colleagues (Taylor, 1981; Taylor et al., 1978; see also Lord & Saenz, 1985) conducted a series of studies demonstrating such effects. As sub­ jects listened to an audiotape of a group discussion, a slide showing the person speaking at any given time was projected. By varying the race of the photos shown, the racial composition of the group could be manipulated experimentally while holding constant the content of the discussion. Thus, subjects in one condition saw three White and three Black males, whereas those in another condition saw five White males and one Black male. In the latter condition, the one Black per­ son is highly salient due to group composition and, in particular, is more salient than when he appeared in the integrated group. The interesting comparison is in perceptions of the same stimulus person (same photo, same voice, same con­ tent) in the two different contexts, only one of which heightens his salience. The results showed that the salient stimulus person drew more attention, was per­ ceived as being more prominent in the group, and was rated more extremely on trait scales than was the same person in the integrated condition. Deviation from White Male Norm. Salience also may be created by member­ ship in certain social groups. Zarate and Smith (1990) argued that groups that deviate from what is perceived as a cultural norm are salient. They argued that in American culture there is a White male norm. That is, it is more “normal” to be White than Black, and more “normal” to be male than female. In terms of race, this effect is created, in part, by the simple numerical preponderance of White people in the country. The historical political dominance of Whites may add to this effect. The same can be said for why males are perceived as more typical Americans than females. Thus, Blacks and women are salient to an extent because they “deviate from” this norm. Zarate and Smith argued that this salience affects categorization processes. For instance, Black males deviate from the White male norm in terms of race but not gender. Thus, Black males are more likely to be categorized as Blacks than as males, because their Blackness is salient. Simi­ larly, White females are more likely to be categorized as females than as Whites. Their gender is salient, whereas their race is not. Zarate and Smith (1990) tested these hypotheses using a category verification task. On each of a series of stimulus trials, subjects were presented a category label (white, black, man, woman) followed by a photograph of a Black male, White male, Black female, or White female. The subjects’ task was to respond “yes” or “no,” depending on whether the person shown in the photograph fit the category label. The speed of responding “yes” to a match between label and pho­ tograph was interpreted as a measure of the subject’s dominant categorization of the stimulus person. For example, subjects should more quickly identify a match

1.

STEREOTYPES

11

between the label “black” and a photograph of a Black person than they identify a match between “white” and a photograph of a White person. Similarly, subjects should be faster in matching photographs of women with the label “female” than they are in identifying that photographs of males fit the “male” label. In general, then, the White male norm hypothesis predicts that males should be categorized more quickly by race, whereas females should be categorized more quickly by gender. In two studies reported by Zarate and Smith (1990), these hypotheses received partial support. As predicted, following race labels, subjects responded more quickly to male than to female photographs; and following gender labels, females were categorized more quickly than males. On the other hand, two general effects were problematic for the hypothesis: Overall, subjects responded more slowly to photographs of Blacks than of Whites, and more slowly to photographs of women than of men. Given that those are the cues by which persons deviate from the White male norm, these slower overall response times are difficult to understand within this framework. Stroessner (1992) tested the White male norm hypothesis using a modifica­ tion of Zarate and Smith’s (1990) procedure. Rather than providing specific category labels, the cues instructed subjects to identify either the race or the gender of the person shown in the photograph. In general, Stroessner’s findings were consistent with predictions. For example, judgments of Black males’ race was fast and of their gender was slow, compared with those of White males (for whom response times for these two identifications were equivalent). Stroessner’s results were particularly informative about the case of Black females, who differ from the perceived norm on both dimensions. When instructed to identify either the race or the gender of stimulus persons, responses to Black females were slow for both dimensions. In a second study, however, Stroessner presented compound category cues (e.g., White male, Black female, etc.) and had subjects identify (yes/no) whether the photograph matched the label. In this case, subjects were able to respond more quickly to Black females than to the other targets. Taken together, these findings support the view that Black females are spontaneously categorized on both race and gender, whereas Black males and White females are categorized spontaneously on only one dimension. Illusory Correlation

We pointed out earlier that any process that leads to perceptual differentiation between groups constitutes a potential basis for the formation of stereotypes. The previous section discussed several ways in which categorization processes can create perceived social units, which then can become the focus of differing belief systems (i.e., stereotypes). Another process that can lead to differential group perceptions is the distinctiveness-based illusory correlation. Hamilton and Gifford (1976) reported evidence that subjects developed differing evaluations of two groups of stimulus persons, even though the two groups were

12

HAMILTON AND SHERMAN

unfamiliar (being identified only as Group A and Group B) and were described by evaluatively equivalent information. These results were due to an illusory corre­ lation (Chapman, 1967) between group membership and behavior desirability, based on the co-occurrence of distinctive (infrequent) stimulus events. Specifi­ cally, in the information presented to the subjects, one of the groups occurred less frequently than the other, making it somewhat distinctive. Similarly, undesir­ able behaviors occurred less often than desirable behaviors, making their occur­ rence distinctive. Due to these differing relative frequencies, when a member of the smaller group performed an undesirable behavior, it constituted the co­ occurrence of distinctive stimulus events. Subjects overestimated the frequency of this category of stimuli, leading them to form less favorable evaluations of the smaller group. Hamilton and Gifford’s (1976) results seemed counterintuitive, because the proportion of desirable and undesirable behaviors describing each group was the same and the normative desirability of the behaviors describing the two groups was carefully matched. Thus, due to something about the way information was utilized, subjects developed differential evaluations of groups that were objec­ tively equivalent. As noted earlier, anything that produces differential percep­ tions of groups can contribute to the formation of stereotypes. Therefore, these results demonstrated that an information processing bias can create such differential perceptions, and hence can lay the groundwork for stereotype formation. Hamilton and Gifford (1976) explained their findings as being due to the dis­ tinctiveness of the co-occurrence of infrequent stimulus information (a member of the smaller group performing an infrequently occurring type of behavior). Given this greater salience, these items become well represented in memory and are easily retrievable when judgments are subsequently called for. Hence, to the ex­ tent that subjects apply the availability heuristic (Tversky & Kahneman, 1973), these items have differential impact on those judgments. Consequently, subjects overestimate the frequency with which members of the smaller group performed undesirable acts, and therefore they make less favorable evaluations of the smaller group. These findings have stimulated a considerable amount of research inves­ tigating these illusory correlations and the mechanisms that produce them (Hamil­ ton & Sherman, 1989; Mullen & Johnson, 1990). The distinctiveness-based explanation advanced by Hamilton and Gifford (1976) assumes that more extensive processing of the distinctive stimuli during encod­ ing makes them more accessible in memory and therefore more likely to be retrieved at the time judgments are made. These judgments are assumed to be at least partially based on the retrieved information (i.e., memory-based judg­ ments). A considerable amount of research evidence is consistent with this in­ terpretation (reviewed by Hamilton & Sherman, 1989; Mullen & Johnson, 1990). For example, the finding that the smaller group is evaluated less favorably than a larger one, even though the two groups are described by evaluatively equiva­ lent information, is quite reliable across a number of studies (see Mullen &

1.

STEREOTYPES

13

Johnson, 1990, for a meta-analysis of these studies). When the infrequent be­ haviors are desirable, the smaller group is evaluated more favorably, rather than less favorably, indicating that the former result is not due simply to group size effects (Hamilton & Gifford, 1976). Also, subjects consistently overestimate the number of undesirable behaviors performed by the smaller group, whereas they typically are quite accurate in their estimates about the majority group. When subjects are asked to indicate the group membership of the person who performed each of the behaviors, they tend to overattribute the undesirable behaviors to the smaller group. When asked to recall the behavioral information, subjects remember a higher proportion of the distinctive behaviors (undesirable behaviors performed by the smaller group; Hamilton, Dugan, & Trolier, 1985). All of these findings are consistent with the argument that the co-occurrence of infrequent stimulus categories has special impact on subsequent judgments. Other results suggest that illusory correlations reflect judgments that are memory-based, being influenced by the greater accessibility of the distinctive items in memory. For example, the typical illusory correlation results just summarized do not occur when the stimulus behaviors describe individuals, rather than mem­ bers of groups (Sanbonmatsu, Sherman, & Hamilton, 1987), presumably because processing information about individual persons invokes on-line integrative processing so that later judgments are not based on retrieval of accessible items. Consistent with this view, Pryor (1986) showed that illusory correlations of groups do not form when subjects are instructed to develop well-integrated impressions of the groups—instructions that also would induce on-line processing. In a simi­ lar vein, several studies have shown that self-relevant motives can induce on-line processing and thereby modify the nature of illusory correlation results (Maass & Schaller, 1991; Sanbonmatsu, Shavitt, Sherman, & Roskos-Ewoldson, 1987; Schaller, 1991; Schaller & Maass, 1989). Although these findings support the distinctiveness interpretation (Hamilton & Sherman, 1989), two alternative explanations for these illusory correlation ef­ fects have been proposed recently, neither of which posits any special processing of the infrequently occurring items. Smith (1991) showed that a computer simu­ lation based on Hintzman’s (1986) memory model was able to produce results similar to those obtained in illusory correlation studies. This model does not as­ sume any special attentional, encoding, or retrieval processes associated with dis­ tinctive stimuli. Rather, judgments are based on activated memory traces of previously learned stimulus items. Because of their differential frequencies of occurrence, the difference between the number of desirable and undesirable be­ haviors performed by Group A is greater than that difference for Group B. There­ fore, comparable rates of retrieval for all four categories of items would lead this difference to be greater for Group A than Group B. If evaluative judgments reflect the difference between those differences, then the typical illusory correla­ tion effects would be obtained. In fact, this is what Smith’s (1991) computer simu­ lation has shown.

14

HAMILTON AND SHERMAN

A second explanation that does not rely on differential encoding of distinctive information (Fiedler, 1991) posits that, because of the differing frequencies of the various stimulus categories and imperfect learning of those frequencies, there is “information loss” that produces “regression” in subjects’ estimates of those fre­ quencies. Hence, subjects overestimate the frequency of the least frequent item type simply due to regression effects in judgment, rather than to any distinctive properties of those items. Other judgments (e.g., evaluative ratings of the groups) are based on those estimated frequencies, producing the typical illusory correla­ tion results. Although different from each other, both the Smith (1991) and the Fied­ ler (1991) interpretations were able to account for many (although not all) of the findings that typically had been viewed as supporting the distinctiveness in­ terpretation. These papers point to mechanisms that, at minimum, may con­ tribute to the occurrence of illusory correlations. The more important and challeng­ ing theoretical question is whether these reinterpretations vitiate the need to as­ sume that the distinctive items receive any differential processing as the information is encoded. To address this question, two issues become of crucial concern: Are distinctive items processed differently than other item types? Subsequently, when judgments are made, are distinctive items more accessible than other item types? Two recent papers provide evidence relevant to these questions. First, Stroess­ ner, Hamilton, and Mackie (1992) measured subjects’ processing times as they read and encoded the items of stimulus information. They showed that, under standard conditions, subjects spent a longer time processing the distinctive items than the other items. Thus, it appears that these items get some kind of additional processing. Moreover, Stroessner et al. showed that induced mood states disrupt the formation of illusory correlations and that they do so by undermining the differential attention directed to these distinctive items. This research provides the most direct evidence for the differential processing hypothesized by the dis­ tinctiveness interpretation. Second, Johnson and Mullen (in press) measured sub­ jects’ response times in judging whether each stimulus behavior had been performed by a member of the larger or the smaller group. They found that sub­ jects responded more quickly to the undesirable behaviors performed by the smaller group, indicating that these behaviors were more accessible in memory than the other three types of behaviors. These findings (Johnson & Mullen, in press; Stroessner et al., 1992) provide direct evidence for the differential processing and, subsequently, the increased accessibility of distinctive information. Therefore, they provide strong support for the distinctiveness interpretation. In addition, however, recent analyses have shown that other mechanisms also may contribute to the formation of illusory correlations under these conditions (Fiedler, 1991; Smith, 1991). An important task for future research will be to determine the conditions under which each of these mechanisms is likely to be influential.

1.

STEREOTYPES

15

STEREOTYPES AS COGNITIVE STRUCTURES

Up to this point, we have attempted to develop the argument that stereotypes are based on the perceived differentiation between social groups, and that both cate­ gorization processes and illusory correlations can contribute significantly to that differentiation. This differentiation does not, in itself, constitute stereotyping, but it does lay the foundation for the development of stereotypes. As the perceiver ac­ quires knowledge and beliefs about a group, and those beliefs become associated with that group, a stereotype of that group becomes established. This stereotype is stored in memory as a cognitive structure and can then influence subsequent per­ ceptions of and behaviors toward that group and its members. Although this state­ ment may seem intuitively obvious, it masks a host of questions about (a) the nature of that cognitive structure and (b) how it produces effects on judgments and be­ haviors. In this section, we discuss a variety of ways that stereotypes, as cogni­ tive structures, have been conceptualized. The next section examines the ways that stereotypes can influence information processing, judgments, and behavior. Early Approaches

For many years, social psychologists simply defined stereotypes as belief systems about social groups and devoted their efforts primarily to identifying and measuring the content of those belief systems (Brigham, 1971; Hamilton, Stroessner, & Dris­ coll, in press). With the resurgence of the cognitive perspective came more focused concerns about what these structures look like, how they are organized, and how knowledge and beliefs about social groups are stored in memory and subsequent­ ly utilized. Initial advances in this direction reflected the implicit personality theory (IPT) approach to social perception. This work attempted to study the structure under­ lying stereotypic belief systems, identifying the complexity and dimensional struc­ ture of the stereotype by applying techniques such as multidimensional scaling and hierarchical clustering analysis (Ashmore, 1981; Ashmore & Del Boca, 1979; Jones & Ashmore, 1973). Although these analyses moved us beyond simple conceptions of a stereotype as a collection of traits associated with a group concept, this approach soon faced the same limitations that ultimately constrained the usefulness of the IPT approach. That is, the research could define and measure the structural organiza­ tion of the stereotype (i.e., its content, its dimensionality, etc.), but the methods employed were not informative about how that structure influenced one’s per­ ceptions of the group, how it biased subsequent information processing, and so on. Cognitive Representations of Groups and Group Members

To further understand these issues, recent research on stereotypes has investigat­ ed how one’s knowledge about groups and group members is represented in memory and how that knowledge is used in making judgments. If stereotypes

16

HAMILTON AND SHERMAN

are cognitive structures, what do those structures look like? Some approaches to this question view stereotypes as generalized conceptions of the prominent at­ tributes of a group, and, as such, stereotypes are represented in memory as ab­ stractions based on previous learning and experience. Other approaches emphasize the role of exemplar information, with less (if any) role afforded to abstract con­ cepts. To investigate these representations, researchers have drawn on and used cognitive models of categorization and judgment, which concern how prior knowledge is used in categorizing and judging individual instances that may or may not be category members. By understanding these processes, we can gain knowledge about the nature of the underlying cognitive representation. In this section, we summarize both abstraction-based and exemplar-based models of representation and their application to social categorization and judg­ ment processes in the domain of stereotyping. We hope to demonstrate that both types of representation are necessary elements for an adequate understanding of stereotypes. Abstraction-Based Representations

Stereotypes traditionally have been conceptualized as abstract generalized be­ liefs about social groups, with little, if any, attention to the role of one’s knowledge about individual group members, or exemplars. In such abstraction-based view­ points, a generalized conception of a group develops as information about that group is acquired. This information may be acquired from a variety of sources, including through first-hand experience with group members and through social learning from other sources, such as family, friends, and the media. As increas­ ing numbers of individual group members are encountered, and as more is learned about the group from other sources, an abstract group representation summariz­ ing this information is formed and stored for future use (Hayes-Roth & HayesRoth, 1977; Posner & Keele, 1968). A number of different terms have been used to refer to these abstract representations, including schemata, prototypes, and knowledge structures. Although distinctions among these terms may be relevant in some contexts (Fiske & Taylor, 1991), for our purposes they can be used in­ terchangeably. In this view, stereotypes are abstract cognitive representations that summarize one’s knowledge about groups in generalized form. That is, they are group prototypes. Any group judgment process must be preceded by an initial act of social categorization. That is, before a group stereotype can be activated and used in one’s perceptions of another person, that target person must be categorized as a member of the group. This process can be influenced by the stereotype. According to “prototype” models of social categorization, categories do not have defining features or criteria that determine whether a person is a member of the category. Rather, categories are “fuzzy sets” whose members vary in the degree to which they fit within the category. Some instances are better examples

1.

STEREOTYPES

17

of the category than others. This “fit” is determined by a comparison of the in­ dividual instance to a category prototype that summarizes the features of category members. If the features of the individual are sufficiently similar to the features of the group stereotype, then the person will be categorized as a member of that group. An individual person can be categorized in terms of several possible categories to which the person belongs. Generally, the target will be categorized according to the group prototype that provides the closest match to the target’s features (Brewer et al., 1981; Cantor & Mischel, 1979; Rosch, 1978). As noted earlier, some writers (Brewer, 1988; Fiske & Neuberg, 1990; Zarate & Smith, 1990) have suggested that categorization is an automatic process based on certain primitive features (e.g., skin color, gender, age) that serve as impor­ tant distinctions and have been used frequently in the past. In this case, the “similar­ ity matching” process involved in categorization relies on a single salient cue, rather than on resemblance to the group stereotype. Then, upon categorization, the stereotype will be activated. This argument suggests that it may be possible to activate a stereotype without having had to use it for categorization. Put another way, a stereotype may be activated and used in subsequent information process­ ing even if it was not used for initial categorization. The relationships among categorization processes, stereotype activation, and stereotype utilization represent important theoretical and empirical issues that recur at several points in this chapter. In summary, abstraction-based models of representation posit that stimuli are categorized by comparison to group prototypes or stereotypes. Stereotype acti­ vation permits the perceiver to infer a target’s features through group member­ ship and can bias subsequent information processing in stereotype confirming ways. Limitations of Abstraction-Based Models

The basic assumptions of abstraction-based representations have been at least implicit in the literature on stereotypes for several decades. Despite their preva­ lence, these models face a number of problems that limit their adequacy as a full accounting of stereotypes. Several of these constraints are summarized briefly in this subsection. Perceivers Possess Group Variability Information. Abstraction-based models of group representation have difficulty accounting for perceivers’ knowledge of and sensitivity to variability in groups. That is, abstraction-based models represent the central tendencies in the perceiver’s beliefs about a target group, but fail to represent one’s perception of the diversity within that group. Yet perceivers readily can judge the degree to which category members vary around the mean on different dimensions. Therefore, it would seem that to judge group variability, perceivers would need to rely on the retrieval of specific category exemplars.

18

HAMILTON AND SHERMAN

However, this issue is not as straightforward as it might seem. If one con­ siders variability within a group to be merely another feature of the group, it is entirely possible that the same processes that govern the acquisition of abstract information about a group’s attributes also could create abstract knowledge about a group’s variability. That is, perceivers could acquire knowledge about group variability in the same way they acquire knowledge about other features (through social learning, through inferences based on observations, etc.). If so, then varia­ bility information could be stored as part of the group stereotype along with in­ formation about other features. This issue has been the subject of considerable debate, with some researchers arguing that variability information is not stored as abstract knowledge (Linville, Fischer, & Salovey, 1989), and others arguing that it is (Park & Hastie, 1987). We return to the issue of perceived variability in a later section. Subtypes. Much of the literature on stereotypes has assumed that perceivers hold and use very generalized conceptions of large social groups (e.g., Blacks, women, Germans, etc.). A potential limitation of such generalized conceptions is that they are overinclusive (i.e., they fail to recognize, at the stereotype level, certain distinctions that perceivers quite obviously make within those broad hu­ man categories). Is a generalized stereotype of women really viable when com­ mon knowledge indicates that this stereotype would not apply equally well to a businesswoman, a barmaid, a homemaker, the checkout clerk at the grocery store, and a female college professor? In recognition of this problem, recent research (Ashmore, 1981; Brewer, 1988; Brewer et al., 1981; Devine & Baker, 1991; also see Deaux, Winton, Crowley, & Lewis, 1985) has shown that such broad stereotypes are often the highest or superordinate categories in a hierarchical category system that also contains more distinct, subordinate levels or subtypes. In fact, it is argued that target persons are typically categorized, and hence stereotypes are created, stored, and used at this subtype level. Brewer et al. (1981) provided a particularly impressive demonstration of stereo­ types that exist at the subtype, as well as the superordinate, level. Subjects sorted photographs of elderly persons into categories on the basis of perceived similari­ ty in personality. There was a substantial degree of consensus among subjects that most stimulus persons fit into three different subcategories. Brewer et al. presented a new sample of subjects with sets of three photographs from the vari­ ous subtypes, and the subjects’ task was to indicate (on an adjective checklist) those attributes that all three of the persons possessed. An attribute was consid­ ered stereotypic of the subtype if at least 50% of the subjects endorsed the attri­ bute as descriptive of all three persons. For each subtype, there were several attributes for which such consensus occurred, and the attribute sets characteristic of each subtype were quite distinct from each other. On the other hand, when

1.

STEREOTYPES

19

subjects were shown three photographs consisting of one person from each subtype (thereby representing the superordinate category), there were very few at­ tributes that subjects agreed described all three of them. These findings indicate that, although people may at times categorize and describe others simply as “old people,” they in fact possess much richer and better differentiated stereotypes of more specific subgroups of that inclusive category. Despite such findings, Smith (1990) questioned the necessity of assuming that stereotypes form at the subtype level. He pointed out there could be infinite num­ bers of subtypes for any category (Black Baptist preacher, Black politician, Black college student, etc.), and many of these subtypes could be context specific (Black running back in Big Ten football games). Smith doubted that perceivers possibly could have stored distinct stereotypes for all of the possible subtypes that they might have developed, and instead argued that many subtype characterizations probably are computed based on the retrieval of exemplars at the time judgments are made. However, Hamilton and Mackie (1990) questioned the necessity of this position. Given extensive (direct or indirect) experience with a group of people, it certain­ ly seems possible that the group representation can become highly differentiated, creating an extremely large number of specific subtypes. Even so, Smith was correct to point out that people can make judgments about groups of people that they had not considered previously as meaningful subgroups (perhaps Black run­ ning backs in Big Ten games). Because these subgroups have no existing represen­ tation, judgments about them necessarily are based on the retrieval of exemplars. The Impact of Instances. Finally, abstraction models have difficulty accounting for the large impact that particular instances can have on processing. For exam­ ple, in one study, subjects expected a newly encountered individual with short hair to be unfriendly, simply because of a previous encounter with an unfriendly individual who also happened to have short hair (Lewicki, 1985). Upon encoun­ tering the second target, subjects retrieved the first individual as a basis of judgment. Although this study is often cited as demonstrating the importance of exem­ plar representations for social judgments, it may not provide strong evidence for nonreliance on abstract knowledge representations, such as stereotypes, because subjects may not have possessed a stereotype relevant to the target in question (i.e., a stereotype regarding the friendliness of long- and short-haired people). Consistent with this view, Smith and Zarate (1990) demonstrated that specific exemplars are utilized as a basis for categorization when abstract knowledge is based on few instances and is weakly defined. Summary. Despite the historical preeminence of generalized stereotypic con­ ceptions, there are good reasons to believe that group categorization and judg­ ment processes do not rely solely on abstract representations of groups. Variability

20

HAMILTON AND SHERMAN

judgments, the role of subtypes, and the impact of specific instances in category judgments suggest that, in addition to generalized stereotypes, people store and use knowledge of individual group members. Next, we turn to a consideration of specific models of exemplar-based processes. Exemplar-Based Representations

Most exemplar-based models were developed to account for categorization processes (Brooks, 1978; Medin & Schaffer, 1978). According to these models, categorization is not achieved by comparing the target to a category prototype. Rather, categorization is achieved by comparing the target to the category mem­ bership of the set of retrieved exemplars. The exemplars retrieved are assumed to be those that are most similar to the target. Exemplar retrieval and usage need not be accessible to consciousness, and an extremely large number of exemplars may be retrieved (Medin & Schaffer, 1978; Nosofsky, 1987; Smith & Zarate, 1990). Because exemplar retrieval is often an implicit process, it may not be revealed by such typical dependent measures as recall and recog­ nition. In their extreme form, exemplar models suggest that stimuli are reacted to on the basis of retrieved exemplars alone, and that categorization per se does not occur. In fact, some models suggest that there is no such thing as abstract, categorical knowledge (Hintzman, 1986). Recently, Smith has extended his exemplar-based model of social categoriza­ tion to cover social judgment processes (Smith, 1990; Smith & Zarate, 1992). According to this extension, not only is exemplar retrieval responsible for the initial categorization of a stimulus, but the retrieved exemplars also guide subse­ quent judgment processes involving the target (Andersen & Cole, 1990; Gilovich, 1981; Lewicki, 1986). Smith (1990) proposed that stereotypes be reconceptual­ ized as summarized exemplars, and not as stored abstract knowledge. Presuma­ bly, the same exemplars that are retrieved for purposes of categorization are utilized to form a group stereotype (Kahneman & Miller, 1986). The features of those exemplars are summarized, creating a stereotype of the particular group into which the target has been categorized, which then acts as an expectancy that can guide subsequent processes. However, to term such a post-hoc group summary a stereotype is to alter the meaning of the term, which has always referred to some form of generalized representation. Moreover, if exemplars are retrieved and summarized for mak­ ing group judgments, the necessity for postulating the creation of a stereotype at all is unclear. Smith’s (1990) conceptualization essentially denies the existence of stored abstract knowledge. However, just as models that deny the existence of exemplar representations have their limitations, so do models that deny the existence of abstract knowledge such as stereotypes.

1.

STEREOTYPES

21

Limitations of Exemplar-Based Models

Social Learning. One basic problem for pure exemplar models is that peo­ ple often acquire abstract information about a group from other sources (e.g., family, friends, television). People may be told that “women are emotional” or that “men are obstinate,” rather than abstracting that knowledge from the retrieval of exemplars. Some exemplar models account for such knowledge by suggesting that ab­ stract information is merely stored and activated as another exemplar when categorization or judgments occur (Hintzman, 1986; Linville et al., 1989; Smith & Zarate, 1990). Thus, when a female target is encountered, perceivers may retrieve the exemplar that “women are emotional” as well as other particular in­ stances of women. Although this conceptualization provides a mechanism for representing abstract characterizations in exemplar form, it blurs the distinc­ tion between abstract and specific information. When such an abstraction is retrieved and used in judgment, how does one know whether it was acquired and stored as a specific instance (as in this example) or is a generalization result­ ing from an abstraction process? This conceptualization lends credence to the view that, as abstraction-based and exemplar-based theories become more richly developed, they approach being indistinguishable from one another (Barsalou, 1990). Exemplars Must Be Held Together by a Theory. To form a coherent group representation, exemplars must be joined together by some sort of category defi­ nition or theory (Medin & Wattenmaker, 1987). There must be some criteria for category inclusion. Smith (1990) noted that, without a category theory, a set of exemplars (e.g., a golden retriever, the number 39, and the graphics board of a computer) does not hang together and form a coherent concept or category. For instance, if someone is asked to describe the typical woman, the exem­ plars retrieved must be constrained by a theory of inclusion criterion. Otherwise, when asked to make category judgments, perceivers could not activate appropri­ ate exemplars. This point is particularly relevant in situations where there is no stimulus other than the category label to act as a retrieval cue. In this case, the category knowledge activated will serve to constrain the set of exemplars that may be retrieved (Kahneman & Miller, 1986; Rothbart & John, 1985). In situations where an actual stimulus target is being categorized or judged, the possibility remains that the target will cue exemplar retrieval and be reacted to without any mediating influence of category knowledge. Note, however, that this view directly contradicts theories of primitive categorization (Bower & Karlin, 1974; Brewer, 1988; Bruner, 1957; Fiske & Neuberg, 1990). Questions regard­ ing the automaticity of social categorization and stereotype activation become of crucial importance for this issue, and are discussed in a later section.

22

HAMILTON AND SHERMAN

Efficiency. Exemplar-based models posit that stereotypic judgments are based on the set of exemplars retrieved at the time the judgment is made, and not on any generalized conception of the target person or group. However, after exten­ sive processing and representation of numerous exemplars (group members), it would seem efficient (even if not necessary) at some point to generalize, to note commonalties among exemplars, and to summarize some of the attained knowledge in the form of group-level characterizations (Hamilton & Mackie, 1990). The alternative—continuing to process, record, and keep track of an ever-increasing number of group members—seems the less efficient, more demanding system. Exemplar models argue that exemplar retrieval is a parallel and implicit process that does not require significant capacity (Nosofsky, 1987; Smith, 1990), so “keep­ ing track” requires neither time nor cognitive resources. Nevertheless, some research has shown that cognitive load can influence the nature of category representations (Rothbart et al., 1978). One troubling aspect of many exemplar models is that they are very difficult to disconfirm. If exemplar retrieval and summation is a parallel, implicit, and unconscious process, it is not clear how one would demonstrate that exemplars have not been retrieved for use. Typical memory measures such as recognition, recall, and response time would not necessarily provide informative data. On-Line Processing. Other analyses indicate that abstraction does seem to occur spontaneously under many circumstances (Fried & Holyoak, 1984; Park & Hastie, 1987; Posner & Keele, 1968). The distinction between on-line and memory-based processes is central to this issue (Hastie & Park, 1986). When perceivers have the explicit goal of forming a group-level judgment as they process information about individual group members, they abstract information from the exemplars and store it (Park & Hastie, 1987). When a group judgment is required, subjects simply retrieve the stored abstraction. However, if subjects have no such processing goal at the time information is encoded, an abstraction may not be formed on-line. In this case, subsequent (memory-based) judgments about the group would require the retrieval and summarization of specific instances. If judgments of a group are memory-based (as posited by exemplar models), one would expect a strong relationship between the exemplar information retrieved and the resulting group judgments. However, some research (Judd & Park, 1988; Park & Hastie, 1987) has found that group judgments can be independent of ex­ emplar retrieval, suggesting that group-level characterizations had been abstracted on-line and used in subsequent judgments. Abstraction Use in Judgments. Regardless of whether a group judgment has been created on-line or through exemplar retrieval, it presumably would be func­ tional to store that judgment for future use. Particularly in cases in which the judgment is likely to be required again and again, it seems highly inefficient to continually recalculate the judgment through exemplar retrieval. Research has

1.

STEREOTYPES

23

demonstrated that pre-formed judgments, and not the original exemplars from which those judgments were derived, are retrieved for subsequent judgment processes (Carlston, 1980; Fazio, Lenn, & Effrein, 1984; Lingle & Ostrom, 1979). Summary. Just as pure abstraction models have difficulties accounting for some findings, so do pure exemplar models. Social learning processes, the neces­ sity of theories to make sense of exemplars, efficiency problems, on-line forma­ tion of group judgments, and the use of pre-formed abstractions in subsequent processing, all pose problems for pure exemplar models. Mixed Models of Group Representation

Given the limitations of both pure abstraction-based and pure exemplar-based models, many researchers have concluded that a viable model of social categori­ zation and judgment processes must include both abstract knowledge representa­ tions, such as stereotypes, and specific group exemplars (Hamilton & Mackie, 1990; Homa, Dunbar, & Nohre, 1991; Kahneman & Miller, 1986; Klein, Loftus, Trafton, & Fuhrman, 1992; Malt, 1989; Murphy & Medin, 1985). Given such a mixed model, a number of important questions arise. One obvious issue is to determine the conditions that lead to relatively exemplar-based or abstractionbased processing. Several relevant considerations can be cited. Learning Order of Information. Based on research by Medin, Altom, and Murphy (1984), Smith and Zarate (1990) designed an experiment to examine the use of abstract and exemplar information in social categorization. The subjects’ task was to learn to classify members of two different groups. Some subjects learned prototype information before they learned the group exemplars; other subjects were given no information about the group prototype. After the learning phase of the experiment, subjects were asked to categorize the learned targets as well as a set of new test targets. The stimuli were designed such that patterns of categorization would differ depending on whether subjects used an exemplarbased or abstraction-based categorization strategy. The results showed that when subjects were given no group prototype information, they categorized new ex­ emplars by comparing them to previously learned exemplars. However, subjects who first learned group prototype information based their categorizations on com­ parison with that prototype. These data suggest that the order in which perceivers learn abstract and exem­ plar information about a group can affect subsequent processing of group mem­ bers. Thus, if perceivers possess a previously formed stereotype when they encounter group members, the stereotype may direct subsequent processing. However, when perceivers interact with members of a group about whom they have no stereotype, retrieved category exemplars may play a particularly impor­ tant role in subsequent processing.

24

HAMILTON AND SHERMAN

On-Line Versus Memory-Based Processing. As noted previously, the extent to which abstract representations are formed as new exemplars are encountered is an important determinant of the kind of information that will be used for sub­ sequent processing. If a group-level characterization has been induced during ex­ emplar encoding, that abstraction should be particularly likely to influence subsequent judgments (Fried & Holyoak, 1984; Hastie & Park, 1986; Homa et al., 1991; Posner & Keele, 1968). However, if no abstraction has been formed on-line, perceivers necessarily will have to rely on exemplar retrieval to guide subsequent processing. An important question for future research concerns the extent to which stereotypes, in the absence of on-line judgment formations, pro­ vide abstract information that precludes the retrieval of exemplars for judgment purposes. Conditions o f Limited Capacity. Cognitive capacity affects the processing of information about group members. As already noted, Rothbart et al. (1978) showed that, under conditions of constrained capacity, perceivers were more likely to organize information in terms of a group than individual members. Other research indicates that cognitive demands affect the likelihood that perceivers will rely on category-level knowledge, rather than individuating target information, when judging a stimulus person (Bechtold, Naccarato, & Zanna, 1986; Bodenhausen & Lichtenstein, 1987; Fiske & Pavelchak, 1986). There also is evidence that stereotypes are spontaneously or automatically activated (Devine, 1989; Macrae, Milne, & Bodenhausen, in press) or, upon activation, automatically applied (Gil­ bert & Hixon, 1991). These findings suggest that the use of abstract information requires little capacity, and that perceivers are more likely to rely on abstract information under capacity constraints. Exemplar-based models propose that issues of cognitive capacity are irrele­ vant to exemplar retrieval and summation (Smith, 1990), because exemplar processing is assumed to be implicit and parallel. Given these assumptions, evi­ dence that capacity limitations can affect exemplar use would be difficult to ob­ tain. Clearly, the effects of capacity load on the differential use of exemplar and abstract knowledge are undetermined at this point. Recent Encounters. As Lewicki (1986) demonstrated, recent encounters can affect subsequent judgments. The strength of these effects is determined by the recency of the encounter and the availability in memory of the exemplar (Kahne­ man & Tversky, 1973). In addition, the exemplar is likely only to affect judg­ ments of targets that bear some similarity to the exemplar (Kahneman & Miller, 1986; Smith, 1990). This type of analogical processing likely decreases the ef­ fects of abstract category knowledge. Summary. A number of factors appear to mediate the extent to which categori­ zation and judgment processes rely on abstract or exemplar information. Whether exemplar or abstract information is learned first affects which type of informa­

1.

STEREOTYPES

25

tion will dominate subsequent processing. Judgments that are formed on-line are unlikely to rely on exemplar retrieval, whereas memory-based judgments rely on exemplar retrieval. Under conditions of limited capacity, subjects may be less likely to rely on exemplar retrieval, and recent encounters can influence the use of abstract and exemplar information. Representation of Group Variability Information

One of the cardinal features of stereotypes, dating back to the earliest writing on the topic (Lippmann, 1922), is that stereotyping represents an overgenerali­ zation in that attributes are ascribed to all group members, leading to perceptions that “they are all alike.” However, it is also clear that perceivers are at least some­ what sensitive to differences among the members of groups. Members of some groups do seem to be all alike or homogeneous, whereas members of other groups appear to be more diverse or heterogeneous. In more general terms, groups differ in their perceived variability. This means that perceivers somehow can assess the degree of variability in those groups. How is this achieved? Is that perception of variability a part of one’s stereotype of a group? Research on perceptions of group variability was stimulated by the highly replicable out-group homogeneity effect (OHE)—the tendency of people to per­ ceive out-groups as being more homogeneous than groups to which they belong (for discussions of this literature, see Brewer, 1993; Mullen & Hu, 1989; Ostrom & Sedikides, 1992; Quattrone, 1986). Although the OHE was the catalyst, the mechanisms underlying perceptions of group variability are also of central importance for understanding stereotypes as mental representations of groups. Perceived group variability has been shown to affect many aspects of stereotyp­ ing and intergroup perception, including (a) the probability that a stereotype will be applied to a particular group member (Park & Hastie, 1987; Park, Judd, & Ryan, 1991), (b) the likelihood of generalization from one group member to the whole group (Park & Hastie, 1987), and (c) the likelihood of generalization from one group member to another (Quattrone & Jones, 1980). Perceived variability affects both social categorization and social judgment processes (Park & Hastie, 1987), and it may have a significant impact on stereotype change (Park et al., 1991). Several of these points are addressed in other sections of this chapter. We focus here on how information pertaining to group variability is represented in memory and how variability judgments are made. Models of Group Variability Representation

Several models have been proposed that offer differing positions on these is­ sues. Some models argue that stereotypes, as group-level representations, can­ not account for perceivers’ sensitivity to the degree of variability among a group’s members. In these accounts, variability information is not represented at the group

26

HAMILTON AND SHERMAN

level, but is derived from exemplar retrieval. Other models hold that variability information is stored at the group level, arguing that variability information is abstracted in the same manner as information about any other attribute and is stored as part of the group stereotype. We summarize four models of perceived variability and discuss their implications for stereotype representation. The PDISTModel As mentioned earlier, one of the reasons some research­ ers have turned to exemplar models of social cognition is to account for perceivers’ sensitivity to variability within groups of people (Linville et al., 1989; Smith, 1990). Exemplar models argue that group-level information cannot account for variability knowledge. Rather, variability estimates are based on retrieved group exemplars. One of the most clearly specified models of variability representation is the exemplar model proposed by Linville and her colleagues, called PDIST (Lin­ ville et al., 1989; Linville, Salovey, & Fischer, 1986). According to the model, judgments of variability are created by retrieving particular group exemplars and summarizing their features. In PDIST, the degree of perceived variability depends primarily on the num­ ber of group members known. If one retrieves a large number of exemplars, those instances are likely to have more variability than when a small number of exem­ plars is retrieved. One consequence is that typically the perceived variability of familiar groups (with many easily retrievable exemplars) will be higher than the perceived variability of unfamiliar groups (with few easily accessible exemplars). According to PDIST, the OHE is a result of differential familiarity with in-groups and out-groups: People simply know more in-group than out-group members, resulting in greater perceived variability in the in-group. The Linville et al. (1989) model does allow for the representation of abstract knowledge (e.g., the statement that “men are obstinate”). However, that “abstract” information about the group is stored, retrieved, and weighted as just another exemplar. In this model, then, group variability information is not stored as part of a group representation. Variability estimates are strictly retrieval based and are cal­ culated only when they are specifically requested. What evidence supports the PDIST model? Computer simulations have demon­ strated that, in a retrieval-based judgment process, group familiarity does affect perceived variability, as predicted (Linville et al., 1989). In addition, Linville et al. (1986, 1989) showed that perceived variability increases as familiarity with groups increases over time, and that variability estimates are equal for groups that are known equally well (males and females). However, the relationship between familiarity and perceived variability is not clear cut. Some researchers have found that greater familiarity does not lead to greater perceived variability (Jones, Wood, & Quattrone, 1981). Others have found in-group/out-group differences in perceived variability between genders, a case

1.

STEREOTYPES

27

where degree of familiarity should be comparable (Park & Rothbart, 1982). Still others have found in-group/out-group differences using a minimal group situa­ tion that equates knowledge of the two groups (Judd & Park, 1988; Mackie, Sher­ man, & Worth, 1993). Finally, Karasawa and Brewer (1992) demonstrated that in some cases greater familiarity leads to lower perceived variability. Clearly, the relationship between familiarity and perceived variability is not as direct as posited by the PDIST model. Dual Predictor Model. A second exemplar model of variability representa­ tion is the dual predictor model recently proposed by Kashima and Kashima (1993). Like PDIST, this model assumes that variability information is not stored with the group stereotype, but is instead derived from the retrieval of particular group members. However, the exemplar summarization process in the dual predictor model is markedly different from the one in PDIST. The dual predictor model assumes that the similarity of two exemplars is de­ termined dually by the number of features the two exemplars share and the num­ ber of distinct features possessed by each exemplar (Tversky, 1977). As the number of shared features increases, similarity increases; as the number of dis­ tinct features increases, similarity decreases. Group variability judgments are es­ sentially judgments of the similarity among group members. When a group variability judgment is called for, group exemplars are retrieved and their similar­ ities and differences are assessed. The group variability estimate is based on the overall numbers of similarities and differences that occur in the retrieved ex­ emplars. Unlike PDIST, variability is not determined by the raw number of exemplars retrieved, but rather by the particular qualities of those exemplars. If all the ex­ emplars retrieved are very similar, the variability estimate will be low, even if there are many of them (Quattrone, 1986). On the other hand, variability judg­ ments may be quite high even if only a few (but highly dissimilar) exemplars are retrieved. In support of their argument, Kashima and Kashima (1993) found that increased similarity information resulted in lower perceived variability, whereas increased difference information resulted in higher perceived variability, as expected. However, they also found a significant main effect due to familiarity: greater familiarity was associated with greater perceived variability, independently of similarity and difference information. Clearly, further research on this model is necessary. In discussing PDIST and the dual predictor model, we have examined the ef­ fects of familiarity and similarity on perceived variability, because of their relevance to these models. However, these data do not directly assess whether variability estimation actually relies on exemplar retrieval, as these models ar­ gue. In fact, there is a paucity of evidence bearing directly on this central ques­ tion. The Linville et al. (1989) computer simulation demonstrated that such a

28

HAMILTON AND SHERMAN

retrieval-based process is sufficient to produce the predicted effects. However, variability estimates may not necessarily be retrieval based. One indication of retrieval-based processes is a relationship between judgments and recalled information. If perceivers retrieve exemplar information to make their judgments, then that information should be positively related to the result­ ing judgment. Some research has failed to find such a relationship. Park and Hastie (1987) found that repeating certain behaviors (presumably making them more memorable) did not affect subsequent variability judgments. Judd and Park (1988) also failed to find a relationship between recalled information and variability judg­ ments. However, the conclusion from these findings that the variability judgments were not retrieval based may be premature. For example, in the Judd and Park (1988) study, the information presented to subjects was visible when the judg­ ments were requested (see Linville et al., 1989), therefore judgments did not need to rely on retrieval at all. In contrast, Mackie et al. (1993) found a relationship between retrieved infor­ mation and variability judgments, suggesting a retrieval-based process. Specifi­ cally, they found a relationship between recalled similarity information and perceived variability, as predicted by the dual predictor model, although not be­ tween recalled difference information and perceived variability. Mackie et al. (1993) also assessed subjects’ response latencies for making variability judgments, and obtained results that were consistent with a retrieval-based process. Although both in-group and out-group judgments appeared to be retrieval based, the in­ group judgments took less time than the out-group judgments. This finding con­ tradicts the PDIST model, in that if more in-group than out-group members are retrieved, judgments of in-group variability should take longer. Given these mixed findings, it is difficult to draw any strong conclusions about the extent to which variability judgments are based on retrieval of exemplars from memory. Moreover, if that process of exemplar retrieval and summation is pre­ sumed to be a parallel and implicit process in exemplar models, it may be difficult to obtain evidence for or against these models. That is, one would not necessari­ ly expect to find a relationship between judgment and recall, nor would response latency data necessarily be informative about a process that is assumed not to consume resources and capacity. The issue of whether variability judgments are formed in a strictly retrieval-based fashion will require further research. Abstraction-Plus-Exemplar Model. In contrast to the exemplar models just discussed, some researchers have argued that our definition of stereotypes must be broadened to include variability as well as central tendency information. Park and Judd (Judd & Park, 1988; Park & Hastie, 1987; Park et al., 1991) proposed that variability estimates are updated on-line and stored as part of the group stereo­ type, just like information about any other attribute (Fried & Holyoak, 1984). When required, the stored variability estimate may be retrieved; exemplar retrieval is unnecessary. Particular exemplars also are stored in memory and can be

1.

STEREOTYPES

29

retrieved to make judgments. However, exemplars are retrieved in addition to, not instead of, pre-stored abstract knowledge. We refer to this model as the abstraction-plus-exemplar model. According to this model, the out-group homogeneity effect results from differ­ ences in the kinds of information used for judging in-group and out-group varia­ bility. Whereas judgments of both groups rely on the retrieval of abstract variability estimates, judgments of the in-group (but not the out-group) also are based on some retrieved exemplars. Use of these exemplars should produce higher esti­ mates of variability for in-groups than out-groups. Park et al. (1991) suggested several reasons why exemplars may play a larger role in in-group than in out-group judgments. First, we simply may be more in­ terested in different kinds of information about in-groups and out-groups (Park & Rothbart, 1982). Specifically, we may be oriented toward defining the typical qualities of out-group members (which focuses on abstract information), but in identifying differences between ourselves and other in-group members (which focuses on in-group exemplar information; Brewer, 1993). Second, the self is more likely to come to mind as an exemplar when making judgments about one’s in-group than about an out-group (Park et al., 1991). This may induce compari­ sons between the self and other in-group members, focusing attention on in-group exemplars both at encoding and retrieval. Third, we may be more motivated to form accurate impressions of in-groups than of out-groups. If so, then we may use a larger sample of exemplars when making variability estimates of an in­ group than an out-group. Finally, we may be exposed to different kinds of infor­ mation about in-groups and out-groups. Our impressions of in-groups are more likely to be based on actual behavioral exemplars that we have witnessed, whereas our knowledge of out-groups often is provided (by media or socializing agents) in the form of general descriptors. We simply may have had less direct experience with individual out-group members (Park et al., 1991; Quattrone, 1986). Most of the evidence cited as support for on-line processing of variability in­ formation comes from failures to obtain relationships between recalled informa­ tion and variability judgments (Judd & Park, 1988; Park & Hastie, 1987). However, as already discussed, these data are not conclusive and other findings (Mackie et al., 1993) argued against on-line variability abstraction. However, there is support for some of the other predictions of the abstractionplus-exemplar model. For example, Park and Judd (1990) obtained positive corre­ lations between self-judgments and in-group judgments, but no relationship be­ tween self-judgments and out-group judgments. Also, Park and Judd reported evidence of a relationship between retrieved in-group members and in-group judg­ ments, but no relationship between retrieved out-group members and out-group judgments. These data all support the model’s contention that retrieval processes play a larger role in in-group than in out-group variability judgments. On the other hand, when the effects of self and retrieved group members were controlled, the OHE still existed, suggesting that other factors contribute to the OHE (Park

30

HAMILTON AND SHERMAN

& Judd, 1990). Also, as noted earlier, the Mackie et al. (1993) finding that in­ group judgments were made more quickly than out-group judgments questions the assumption of greater in-group than out-group exemplar retrieval. This model proposes several possible mechanisms that may contribute to judg­ ments of group variability. To date, there is little evidence that variability infor­ mation is stored as an abstract representation as part of the group stereotype, questioning one of the model’s major contentions. Even if abstract information plays little or no role, the model suggests several self-involving and motivational factors that also may influence variability judgments. Frequency Distribution Model. Another model has been proposed by Park and Judd (Kraus, Ryan, Judd, Hastie, & Park, 1993; Park, Ryan, & Judd, 1992). The frequency distribution model posits that people spontaneously create mental frequency distributions that summarize the number of group members at differ­ ent levels of particular attribute dimensions. For instance, subjects may store the number of high-, moderate-, and low-intelligence behaviors performed by group members, or the number of smart, average, or stupid individuals in the group. These frequency distributions themselves do not constitute variability estimates. Rather, when a variability judgment is required, subjects retrieve the distribu­ tions and base their variability estimate on the number of levels, or subtypes, used to discriminate among group members along particular dimensions. Thus, a group represented by five different levels (or subtypes) of intelligence will be judged as more variable than a group represented by three levels. In this view, the out-group homogeneity effect occurs because in-group members are more likely to be subtyped than out-group members, due to the same processes differen­ tiating self from other in-group members outlined earlier. In contrast to the abstraction-plus-exemplar model, this model does not postu­ late that stereotypes include both variability and central tendency information. The formation of distributions occurs on-line, but variability estimates are based on the retrieval of subtypes. This retrieval process is analogous to PDIST, but rather than focusing on the number of exemplars retrieved, the frequency distri­ bution model focuses on the number of subtypes retrieved. In contrast to PDIST, variability information is attended to on-line, but (in line with PDIST) no ab­ stract variability representation is stored at the group level. Interestingly, this model portrays a different role of subtyping in stereotype maintenance than typically has been assumed. Traditionally, subtyping has been viewed as protecting stereotypes from change; inconsistent group members are subtyped, leaving the overall group impression intact. However, in this model, the subtyping process may yield greater perceived variability within the group, and therefore undermine stereotyping. This analysis proposes an interesting para­ dox that is addressed further in the section on stereotype change. Although only recently introduced, some initial support for the frequency dis­ tribution hypothesis has been reported (Kraus et al., 1993; Park et al., 1992).

1.

STEREOTYPES

31

As predicted, subjects generated more in-group than out-group subtypes, and varia­ bility judgments were positively related to the number of subtypes used to classi­ fy group members. However, the judgment process hypothesized by the model—that people retrieve subtypes when they make variability judgments— remains untested. Summary. Attempts to understand how perceivers understand and make judg­ ments of group variability obviously have produced a growing research litera­ ture, and there has been more than a little debate among advocates of the various models we have discussed. In large part, this research has been driven by the continuing puzzle of the out-group homogeneity effect: Why is it that people often perceive greater homogeneity in other groups than in their own? Intuitively the finding is not surprising, but developing a fully satisfactory explanation has not been easy (a task further complicated by recent findings of greater perceived in­ group homogeneity; Brewer, 1993; Simon & Brown, 1987; Simon & Hamilton, in press). But the fundamental issue is considerably more than the out-group homogeneity effect; the issue concerns how information about group members is processed, stored in memory, and used in making judgments about the group. That is, it leads us to think about the nature of group representations and what we mean by the term stereotype. As stated earlier, a stereotype pertains to “central tendency beliefs,” that is, the attributes thought to characterize a group “as a whole” or “on average.” But people certainly retain knowledge of individual group members and of their ex­ periences with them, and perceivers can make at least reasonable estimates of group variability. Moreover, perceived variability is related to a number of ef­ fects in stereotyping and intergroup perception (Hastie & Park, 1987; Park et al., 1991) and intergroup behavior (Wilder, 1978). What role, then, do percep­ tions of variability play in stereotyping? At present, there is little evidence for on-line abstraction of variability infor­ mation. However, if variability estimates are generated in a strictly retrieval-based fashion, it becomes unclear what causal role perceptions of variability could play in producing such effects. That is, when performing many of the tasks often related to variability judgments (e.g., generalization tasks), it seems highly unlikely that subjects would spontaneously retrieve information and form a variability esti­ mate prior to responding. If they do not, it is questionable that post hoc, retrievalbased variability estimates would play a causal or mediating role in producing other, related effects. Instead, some other aspect of group information, such as stereotypes or central tendency information, simply could exert direct influence on these responses. That is, it seems possible that variability estimates are mere­ ly by-products of stereotyping, rather than contributors to it. Is such a view tenable? Perhaps variability estimates simply reflect the strength with which a stereotype is held. For example, the more confidence or convic­ tion one has in a stereotype, the more that stereotype might be used in making

32

HAMILTON AND SHERMAN

judgments of the group, with relatively little reliance on retrieval of exemplar information from memory. In this case, central tendency judgments could be made with relative ease, and the group would be seen as rather homogeneous. On the other hand, for less clearly articulated and established stereotypes, subsequent judgments may rely more heavily on retrieval of exemplars and/or subtypes, which typically might increase perceptions of variability. Could this view account for the out-group homogeneity effect? Perhaps. As­ sume that perceivers develop less clearly defined stereotypes of their own group, and hence those self-stereotypes are held with less conviction. In contrast, stereo­ types of other groups are formed more readily and are held with greater confi­ dence. Given this difference, judgments of one’s own group would draw on relevant exemplar information, whereas judgments of other groups would be based more directly on the stereotype. It would follow, then, that estimates of in-group variability would be greater than would those for out-groups. This view is also compatible with recent research on self-stereotyping, in which both greater selfstereotyping and perceived in-group homogeneity have been found under condi­ tions that would be likely to promote clear and confident self-stereotypes, such as minority status and heightened salience of in-group (Brewer, 1993; Simon & Hamilton, in press). If variability estimates derive from stereotypes and if other variability-related effects are also products of stereotypes, there certainly would be a positive rela­ tionship between these effects and perceived variability. However, to the extent that variability estimates are retrieval based and are made only when they are explicitly requested, we must begin to question the causal or mediational role of perceived variability in producing these effects. This discussion highlights the importance of understanding the basic represen­ tational structure of group-relevant information. Earlier we argued against a pure exemplar-based definition of stereotypes. We questioned why, if categorization and judgment processes can be based exclusively on exemplar information, one would construct a retrieval-based stereotype at all. We summarized evidence that, we argued, supports a conception of stereotypes as abstract representations of, and derived from, group-relevant information. However, if group-relevant in­ formation is represented in generalized form, why is variability information not abstracted and stored as part of the stereotype? The answer may be in the relative importance that the perceiver attaches to variability information. That is, the fact that we develop abstract knowledge about some aspects of group-relevant infor­ mation does not mean that all aspects will necessarily be represented abstractly. Thus, those attributes that are most important for defining a group, or that are most useful for perceivers in making judgments and guiding behavior, are more likely to be abstracted and represented in a stereotype. At this point, it appears that variability information may not be one of those central attributes. Clearly, there is more work to be done on these issues. The models proposed thus far, and the research they have stimulated, have left many questions un­

1.

STEREOTYPES

33

answered. Nevertheless, they have challenged traditional ways of thinking about the nature of stereotypes, and they have generated new conceptual frameworks that are likely to be the catalyst for additional progress in the years ahead.

STEREOTYPING AND INFORMATION PROCESSING

As mentioned earlier, stereotypes act as expectancies about groups and their mem­ bers (Hamilton et al., 1990). Like all expectancies, stereotypes guide informa­ tion processing and often are perpetuated by confirmatory biases that they themselves generate. Indeed, the activation of a stereotype can affect all aspects of social information processing, including attention (Zadny & Gerard, 1974; Bodenhausen, 1988), interpretation (Darley & Gross, 1983; Sagar & Schofield, 1980), inference (Bodenhausen & Wyer, 1985; Krueger & Rothbart, 1988), and retrieval (Bodenhausen & Lichtenstein, 1987; Cohen, 1981; Hamilton & Rose, 1980; Stangor & Duan, 1991). Stereotypes also can influence the types of infor­ mation perceivers seek about targets (Kunda, 1990; Skov & Sherman, 1986; Snyder & Swann, 1978) and can direct behavior in confirmatory ways, creating self-fulfilling hypotheses (Snyder, Tanke, & Berscheid, 1977; Word, Zanna, & Cooper, 1974). In this section, we focus primarily on how stereotypes affect encoding, retrieval, and inference processes. We place particular emphasis on topics that have received substantial attention in recent years. Stereotypes: Encoding and Retrieval Biases

Interpretation Effects. Once activated, stereotypes can affect the interpreta­ tion of subsequently presented target information (Darley & Gross, 1983; Dun­ can, 1976; Kunda & Sherman-Williams, 1992; Sagar & Schofield, 1980). For instance, Sagar and Schofield (1980) found that ambiguous acts were interpreted as more aggressive when performed by a Black target than by a White target. This type of bias is particularly strong when behavioral information is ambigu­ ous. Some researchers have suggested that such interpretational biases occur au­ tomatically (Devine, 1989). In a similar vein, Biernat, Manis, and Nelson (1991) showed that people use different standards of comparison when judging the behaviors of different groups of people. For example, an assertive behavior performed by a woman may be perceived as being more assertive than when that same behavior is performed by a man. Manis et al. argued that the assertiveness of the target person’s be­ havior is evaluated in comparison with the standard of typical assertiveness of the person’s gender group, based on stereotypic expectancies. Because peo­ ple believe that women typically are less assertive, a woman’s clearly assertive

34

HAMILTON AND SHERMAN

behavior is viewed as more assertive than when it is performed by a man, for whom the same behavior is less discrepant from his group standard. Another interpretation bias involves the attributions perceivers make for stereotype-consistent and stereotype-inconsistent behaviors. Behavior that con­ firms a stereotype is more likely to be attributed to a target’s stable personality factors than disconfirming behavior (Bodenhausen & Wyer, 1985; Macrae & Shepherd, 1989), whereas inconsistent behavior is more likely to be attributed to situational factors. Furthermore, perceivers differentiate between high- and low-credibility sources only when they provide stereotype-inconsistent informa­ tion (Macrae, Shepherd, & Milne, 1992). There is no preference for highcredibility sources when stereotype-consistent information is provided. Subjects have a more restrictive acceptance threshold for inconsistent information. Stereotypes also can affect the patterns that are perceived in available infor­ mation. Hamilton and Rose (1980) showed that subjects overestimated the num­ ber of times that stereotypic traits were used to describe members of occupational groups. The groups were described with equal numbers of stereotypic and nonstereotypic terms, so there was no actual relationship between the groups and the terms used to describe them. However, subjects perceived illusory correla­ tions between the groups and their stereotypic traits. In another study in which traits did describe some groups more often than others, subjects were more like­ ly to detect the relationship if the trait-group association was stereotypic than if it was not (Hamilton & Rose, 1980). Thus, stereotypes not only influence in­ terpretation of individual behaviors but also affect the associations that are per­ ceived in patterns of acquired information. Selective Processing and Recall. Taylor and Crocker (1981) proposed that schemata such as stereotypes function to filter out stereotype-inconsistent infor­ mation. Accordingly, stereotype-confirming information would receive more at­ tention than inconsistent information (Bodenhausen, 1988; Cohen, 1981; Zadny & Gerard, 1974). Furthermore, this increased attention would cause consistent information to be better incorporated into the perceiver’s impression of the tar­ get. These processes would yield greater recall of stereotype-consistent informa­ tion and would serve to perpetuate the stereotype. Early research suggested a memory advantage for consistent information over stereotype-irrelevant information (Cohen, 1981; Rothbart, Evans, & Fulero, 1979). However, recent analyses of the accumulated research findings showed little evidence of increased encoding or recall of consistent information over in­ consistent information (Rojahn & Pettigrew, 1992; Srull & Wyer, 1989; Stangor & McMillan, 1992). On the contrary, most research has shown an advantage for inconsistent information. However, not all research has been consistent with these general conclusions. Both Stangor and McMillan (1992) and Rojahn and Pettigrew (1992) published extensive meta-analyses of the existing research on memory for expectancy-

1.

STEREOTYPES

35

congruent and expectancy-incongruent information. Their analyses indicate that, although it is true that there is generally a recall advantage for inconsistent infor­ mation, this tendency is modified by a number of important variables. The recall advantage of inconsistent information is due to the increased atten­ tion it receives at encoding. Because it violates an expectancy, it is surprising, draws people’s attention, and initiates attempts to explain the inconsistency (Clary & Tesser, 1983; Hastie, 1984; Sherman & Hamilton, in press; Stern, Marrs, Millar, & Cole, 1984). In general, the meta-analyses demonstrate that any varia­ ble that decreases such inconsistency resolution processes will attenuate recall of inconsistent information, in some cases leading to recall advantages for con­ sistent information. Studies finding better recall of inconsistent information typically have created trait expectancies in the laboratory, whereas experiments using preexisting stereo­ types as expectancies frequently have found better recall for consistent informa­ tion. There are obviously important differences between trait expectancies and stereotypes. Expectancies pertaining to groups seem to generate less of an incon­ sistency effect than expectancies pertaining to individuals (Srull, Lichtenstein, & Rothbart, 1985; Stangor & Ruble, 1989; Stern et al., 1984), perhaps because (compared with stereotypes) trait expectancies are stronger and hence yield greater attempts at inconsistency resolution (Srull et al., 1985; but see Stangor & McMil­ lan, 1992, for a different view). In addition, specific traits certainly circumscribe a narrower range of behaviors than stereotypes (Andersen & Klatzky, 1987), leav­ ing less room for inconsistency. Using preexisting stereotypes, Bodenhausen (1988) showed that subjects had better recall for consistent than inconsistent information. Furthermore, this recall advantage apparently was due to increased attention to consistent information, and not due to interpretational biases or retrieval effects (see also Bodenhausen & Lichtenstein, 1987). There are other variables that eliminate or even reverse the recall advantage of inconsistent information. Subjects with limited cognitive capacity are less able to engage in inconsistency resolution, and therefore recall consistent information at least as well as inconsistent information (Macrae, Hewstone, & Griffiths, 1993; Srull, 1981; Srull et al., 1985; Stangor & Duan, 1991). In fact, several researchers suggested that, as capacity saving devices, stereo­ types are most likely to bias encoding processes under capacity constraints (Boden­ hausen & Lichtenstein, 1987; Macrae et al., 1993; Stangor & Duan, 1991). Under such conditions, the relative ease with which stereotype-consistent information is processed, and the inability to devote appropriate resources to explaining the inconsistent information, will lead to a recall advantage for consistent informa­ tion. In addition, due to limited capacity, subjects may be unable to form judg­ ments on-line. If so, then not only would the memory load yield increased recall for consistent information, the recalled information also may play an important role in judgments of the target.

36

HAMILTON AND SHERMAN

Several experiments provide evidence relevant to these points. Using an ex­ perimentally induced expectancy, Stangor and Duan (1991) demonstrated that the recall advantage of inconsistent information decreased as the complexity of the processing task increased, presumably depleting capacity (see also Hamil­ ton, Driscoll, & Worth, 1989). In a second study, Stangor and Duan obtained a recall advantage for consistent information under limited capacity conditions, but did not find a relationship between the information recalled and the judg­ ments made about the targets. Using expectancies based on occupational stereotypes, Macrae et al. (1993) found better memory for inconsistent information under high-capacity conditions and better memory for consistent information under low-capacity conditions. As predicted, they found a relationship between recall and judgment only under limited capacity conditions. Similarly, Bodenhausen and Lichtenstein (1987) showed a recall advantage for consistent information under anticipated high load conditions. However, like Stangor and Duan (1991), they also found no relationship between recall and judgments. Summarizing to this point, inconsistent information generally is recalled bet­ ter than consistent information, but certain conditions favor recall of consistent information. This latter pattern of results appears to be more likely when the ex­ pectancy pertains to a group (e.g., a stereotype) than when it pertains to an in­ dividual, and when capacity is limited (although it appears that recalled information may not affect judgments). Most researchers attribute this pattern to subjects’ decreased ability to process inconsistent information (Macrae et al., 1993; Srull, 1981; Srull et al., 1985; Stangor & Duan, 1991). Without increased attention during encoding, inconsistent information is recalled less well than consistent in­ formation. To this point, only Bodenhausen (1988) has demonstrated that consistency bi­ ases may involve more than equalizing the attention paid to consistent and incon­ sistent information. He provided evidence that consistent information actually receives more attention than inconsistent information. Given that there is frequently a lack of a relationship between judgments and recall (e.g., Bodenhausen & Lichtenstein, 1987; Stangor & Duan, 1991), Bodenhausen’s findings may be par­ ticularly important. Stereotype-driven biases that favor the recall of consistent information may seem to be of little consequence if that information does not affect subsequent judgments. However, if such biases affect the attention given to consistent information, they will affect both on-line judgments (by affecting the weight given to different information) and recall of information (well-attended information tends to be recalled better). Thus, establishing such stereotypecongruent attentional biases has important implications. In addition, there are other reasons that inconsistent information may have little influence on judgments. Inconsistent information often is explained away or attributed to situational causes as it is encoded. If so, then although this incongruency-resolution process would lead to a recall advantage for incon­

1.

STEREOTYPES

37

sistent information, it also would decrease the impact of that information on sub­ sequent judgments. On the other hand, if subjects are unable to process inconsis­ tent information in this way at encoding (e.g., under a high memory load), they are less likely to form on-line judgments and inconsistent information is unlikely to be well retrieved. Consequently, subsequent judgments would be memorybased, but would not be influenced heavily by the recall of inconsistent informa­ tion. In both of these cases, then, the inconsistent information would not have strong impact on judgments. Response Biases. In their meta-analyses, both Stangor and McMillan (1992) and Rojahn and Pettigrew (1992) found a strong response bias for consistent in­ formation. That is, recognition measures that are not corrected for guessing demon­ strate a strong advantage for consistent information. This suggests that subjects frequently guess that they have seen expectancy-consistent information when they have not. Such strategies may override any effects due to increased encoding and recall of inconsistent information. In a similar vein, Slusher and Anderson (1987) demonstrated that, with the passage of time, it becomes difficult for subjects to distinguish between stereotypeconsistent information that actually was presented and stereotype-consistent in­ ferences that subjects drew from that information. Subjects sometimes are un­ able to distinguish what they know from what they believe. Summary. Stereotypes can serve to bias information processing in a number of self-perpetuating ways. Stereotypes can bias the initial interpretation of infor­ mation in a confirmatory fashion. Attributions made about consistent and incon­ sistent behaviors also serve to confirm the stereotype. Illusory correlations between stereotypical information and targets inflate the perceived prevalence of consis­ tent information. Under appropriate conditions, stereotype-consistent informa­ tion may receive more attention than and be better recalled than inconsistent information. Furthermore, stereotype-consistent information may be recognized falsely as having been encountered, yielding a general response bias for consis­ tent information. In total, these various processes demonstrate that one of the primary functions of stereotypes is self-maintenance. Stereotyping and Inference

One important effect of group stereotypes is that they color our perceptions of individual group members. The stereotype allows us to infer an individual’s charac­ teristics based on group membership. Therefore, stereotypes should reduce the impact of group members’ personal behaviors and attributes in our impressions of them (Bodenhausen & Wyer, 1985; Krueger & Rothbart, 1988). However, this is not invariably the case. Some circumstances favor the use of individuating information over stereotypes. In their now-famous studies, Locksley,

38

HAMILTON AND SHERMAN

Borgida, Brekke, and Hepburn (1980) demonstrated that judgments about the as­ sertiveness of male and female target persons were based on the nature of the behaviors they performed, and not on their gender. This was true even when sub­ jects learned only a single diagnostic behavior about the target. Only when no other diagnostic information was available did gender stereotypes affect subjects’ ratings of the targets. This led Locksley et al. to suggest that stereotypes only will be utilized when the perceiver has no individuating information about the target person. More recent research indicates that this conclusion was somewhat overstated. In a series of studies, Krueger and Rothbart (1988) demonstrated that individua­ tion depends on both the diagnosticity of the stereotype and the diagnosticity of the individuating information. In their first experiment, subjects read only a sin­ gle individuating behavior. In contrast to the Locksley et al. results, even when the behavior was highly diagnostic, stereotypes exerted strong influence on judg­ ments. However, in subsequent experiments, stable behavioral and trait infor­ mation did override stereotype effects. The key to overriding the stereotypes was the presence of temporal consistency in targets’ behaviors. Such consistency is far more diagnostic than a single behavior. Other researchers have documented that individuation occurs only when highly diagnostic information about individuals is provided (Heilman, 1984; Kunda & Sherman-Williams, 1992; Krueger & Roth­ bart, 1988). How, then, do we explain the findings from the Locksley et al. (1980) studies? Krueger and Rothbart (1988) argued that the circumstances in these studies were just right for individuation to occur, in that the behavioral information was highly diagnostic and gender stereotypes may be weak predictors of assertive be­ havior (the target behavior used by Locksley et al.). An experiment by Darley and Gross (1983) demonstrates what may happen when stereotypes are activated in the presence of ambiguous information. Interest­ ingly, whereas Locksley et al. (1980) found that stereotyping occurred only in the absence of specific information, in Darley and Gross’s study, stereotyping occurred only when subjects received specific target information. Subjects who learned about a target person’s socioeconomic status (SES) and received no be­ havioral information did not use their SES stereotype in evaluating the person’s academic ability. However, subjects who received behavioral information in ad­ dition to SES information did differentiate high and low SES students. Analyses revealed that these subjects processed the behavioral information in a hypothesisconfirming manner. Specifically, behavior that was consistent with subjects’ ex­ pectancies was recalled better and weighted more heavily, and ambiguous be­ haviors were interpreted differently depending on the target’s SES. Similarly, there was some evidence from Locksley et al. (1980) that subjects interpreted behavior in a biased manner. Assertive behaviors were rated as more masculine than feminine, and passive behaviors were rated as more feminine than masculine. Whereas male and female targets may have been rated as equally as­ sertive, it is not clear that subjects perceived the targets’ behavior to be equally

1.

STEREOTYPES

39

masculine or feminine. As noted earlier, Biernat et al. (1991) reinterpreted these results by proposing that people use different standards of comparison when judg­ ing males and females. The findings of Locksley et al. (1980) and Krueger and Rothbart (1988) sug­ gest that stereotypes can be overcome if individuating information is strongly di­ agnostic. Similarly, other models suggest that if individuating information clearly contradicts a stereotype, then individuation may occur (Brewer, 1988; Fiske & Neuberg, 1990). Most current theorists predict that stereotypes have clear priority over individu­ ating information in people’s impressions of others (Bodenhausen & Wyer, 1985; Brewer, 1988; Devine, 1989; Dovidio, Evans, & Tyler, 1986; Fiske & Neu­ berg, 1990). This prediction is based, in part, on the simple fact that categorybased stereotypes are more salient and easier to utilize than individuating infor­ mation. Because of capacity requirements, these models suggest that individuation will only occur if perceivers are motivated to attend to target persons carefully. For instance, if perceivers are motivated by personal relevance or desire accura­ cy, they may individuate target persons (Brewer, 1988; Devine, 1989; Fiske & Neuberg, 1990). Also, evidence shows that forming an individuated impression (Fiske, Neuberg, Beattie, & Milberg, 1987) takes longer and requires greater attentional resources (Bechtold et al., 1986) than does a stereotype-based impres­ sion. Stereotypes as Capacity-Conserving Devices. If it is true that stereotypes func­ tion to preserve cognitive resources, then they should be particularly useful in conditions of limited capacity. Recently, a number of studies examined stereo­ type use under high- and low-capacity conditions. In one study (Bodenhausen & Lichtenstein, 1987), when subjects performed a cognitively demanding judg­ ment task, they produced stereotypical judgments and a recall advantage for stereotype-consistent information. This was not the case when they performed a relatively simple judgment task. Macrae et al. (in press) provided stronger support for a capacity-saving func­ tion for stereotypes. Subjects were asked to form impressions of target individu­ als while performing an audiotape-monitoring distractor task. To examine the capacity-preserving function of stereotypes, some subjects were given stereotype labels (e.g., skinhead) along with the information about the target. Presumably, this information would simplify the impression formation task by providing sub­ jects with a stereotypic theme to guide their impressions. The results showed that subjects who were given the stereotype information recalled more of the traits used to describe the target than did subjects who were not given the stereotype. However, this was only true for stereotype-consistent traits. Thus, under capacity constraints, stereotypes function as organizational structures, facilitating the learning of stereotype-consistent information. These subjects also performed better on a test of the material presented in the tape-

40

HAMILTON AND SHERMAN

recorded distractor task. Again, this suggests that the stereotypes functioned to preserve capacity required for the impression task, such that more resources could be directed toward the tape-monitoring task. Macrae et al. (in press) also report­ ed additional experiments confirming this interpretation. Gilbert and Hixon (1991) also showed greater reliance on stereotypes when making judgments under high cognitive load. Additional support for viewing stereotypes as capacity-preserving devices comes from some ingenious studies of the use of stereotypes depending on sub­ jects’ circadian arousal levels (Bodenhausen, 1990). People go through daily var­ iations in their arousal levels that affect capacity and efficiency of working memory. Some people reach their functional peak during the morning (“morning people”), and some in the evening (“night people”). Bodenhausen showed that subjects relied more on stereotypes when they were at the low end of their circa­ dian cycles. Morning people used stereotypes more in the evening and night peo­ ple used them more in the morning. This finding suggests that stereotypes are utilized as capacity-saving devices when people are incapable (or unwilling) to process optimally. Together, the results of these experiments provide solid evidence that stereo­ types function to preserve cognitive resources. The Automaticlty of Stereotyping

Some researchers have argued that stereotyping is not only easier than individua­ tion, but that it is automatic as well (Brewer, 1988; Devine, 1989; Fiske & Neu­ berg, 1990). Why might stereotypes be activated automatically upon categorization? One reason is that many stereotypes apparently are learned at a very early age (Katz, 1976). As a consequence, stereotypes have a long history of activation and are likely to be highly accessible (Higgins & King, 1981). In addition, due to their long-term usage, stereotypes are likely to become some­ what proceduralized (Smith, 1990). Several recent studies have examined the automaticity of stereotyping activa­ tion and application. For example, Devine (1989) demonstrated that nonconscious priming of category labels (e.g., Blacks, Negroes) and terms related to the stereo­ type of Blacks {poor, slavery, athletic) affected subsequent ratings of a raceunspecified target. Devine argued that these results were due to the activation of the Black stereotype in response to the primes. However, it is somewhat un­ clear whether her results were due to stereotype activation or to simple semantic priming (Bargh & Pietromonaco, 1982). Macrae et al. (in press) modified their study on the capacity-preserving func­ tions of stereotypes (described earlier) to test whether these effects occurred au­ tomatically. Subjects performed both an impression task and a tape monitoring task at the same time. Again, some subjects received stereotype labels to assist them in the impression formation task. However, in this study, the stereotype labels

1.

STEREOTYPES

41

were presented subliminally, outside of subjects’ conscious awareness. Their results showed that subjects who were exposed subconsciously to the stereotype labels both recalled more of the traits and performed better on the test regarding the recorded material. These results support the idea that the stereotypes allowed these subjects to conserve resources on the impression task that could then be devoted to the tape monitoring task. They also show that this capacity-preserving func­ tion of stereotypes is not dependent on conscious access to the stereotypes. The stereotypes apparently were activated and used automatically, without conscious awareness. Banaji and Greenwald (in press) presented further evidence that stereotypes can influence perceptions and judgments without intention or awareness. They modified an experiment performed by Jacoby, Kelley, Brown, and Jasechko (1989) demonstrating unconscious influences on memory. Jacoby et al. had found that subjects judged names that they previously had seen (but could not remember seeing) as famous. Those names apparently were somewhat familiar to subjects, even though they could not explicitly remember them, leading to their becoming “famous overnight.” Banaji and Greenwald modified this procedure by making half of the names male and half of the names female. Subjects’ tendency forjudg­ ing previously seen (but unremembered) names to be famous was significantly greater when those names were male than when they were female, suggesting an implicit stereotype that associates males more closely with fame than females. Research has shown that the presentation of trait information in one context can influence subsequent judgments of an ambiguously described target in an un­ related context (Higgins, Rholes, & Jones, 1977; Srull & Wyer, 1979). Banaji and Greenwald (in press) also showed that the effect of a prime depended on its stereotypicality for the judged target. For example, when the trait “dependent” (stereotypic for females, counterstereotypic for males) was used as a prime, sub­ jects subsequently judged a female target person as more dependent and a male target person as less dependent, compared with subjects exposed to a neutral prime. Thus, the stereotypicality of the match between the prime and the target can af­ fect the extent and nature of priming effects. These are implicit, unconscious ef­ fects because they occur without subjects being aware of the influence of the prime. Together, the results of these studies suggest that gender stereotypes can affect information processing and judgments subconsciously. Contrary to these research findings, Gilbert and Hixon (1991) reported results suggesting that stereotypes are not activated automatically. Subjects under a low cognitive load, when exposed to an Asian target person, completed word frag­ ments in a stereotype-biased fashion, suggesting that seeing the target person had activated the stereotype. However, subjects under a high cognitive load did not produce stereotypic word-fragment completions. Because the cognitive load manipulation affected performance in this way, stereotype activation apparently was not automatic, but was subject to capacity limitations. In this study, subjects under high cognitive load in the activation stage were

42

HAMILTON AND SHERMAN

able to identify the target person’s race correctly. Therefore, although stereotype activation may not have occurred in this condition, it was not because subjects failed to categorize the target person. This is a potentially important point. Some models assume that categorization is an automatic process based on perceptible features (Brewer, 1988; Devine, 1989; Fiske & Neuberg, 1990) and that, upon categorization, stereotypes are activated automatically. However, Gilbert and Hixon’s results suggest that categorization can occur without stereotype activation. Summary. Although the findings are not entirely consistent, evidence is ac­ cumulating that stereotypes are activated and applied routinely under some con­ ditions. Stereotypes appear to serve a capacity-preserving function. They are particularly likely to be used under low capacity (Bodenhausen, 1990; Gilbert & Hixon, 1992; Macrae et al., in press) or low motivation (Neuberg & Fiske, 1987). On the other hand, if subjects have sufficient capacity and motivation, and if individuating information is particularly diagnostic or disconfirmatory, stereotypes may be overridden. Just as stereotyping requires less effort than individuation, categorization may require less effort than stereotyping. Categorization appears to be a necessary, but not a sufficient, precondition for stereotype activation and use (Gilbert & Hix­ on, 1992).

AFFECT, COGNITION, AND STEREOTYPING

If there is one domain of social perception where we might expect affective processes to play a particularly important role, it would be intergroup percep­ tion. Historically, emotion and feelings were viewed as major elements contribut­ ing to stereotype-based biases and to the tenacity with which those stereotypes are held (Allport, 1954; Katz, 1981; Lippmann, 1922). Yet, for over a decade, beginning in the mid-1970s, stereotype research was almost totally focused either on the cognitive underpinnings and functioning of stereotypes or on the role of social structures and mass media in the development and maintenance of these belief systems. Discussions of the stereotype literature during this period (Hamil­ ton, 1979; Hamilton & Trolier, 1986; Hewstone, 1989; Stephan, 1985; Stroebe & Insko, 1989) included relatively little coverage of the role of affective factors in stereotyping and intergroup perceptions. However, there were occasional calls for greater attention to affective processes and their interrelations with cognitive processes in this domain (Fiske, 1982; Hamilton, 1981). Fortunately, in recent years, we have witnessed not only a resurgence of in­ terest in affect, emotion, motivation, and related processes per se, but also (and more importantly) a focus on the interactive influence of affect and cognition on social information processing and judgment processes. This development has im­ pacted research on intergroup perceptions and has illuminated a variety of ways

1.

STEREOTYPES

43

in which affect and cognition interface in intergroup perceptions. In this section, we highlight some of these recent developments. A broader and more thorough coverage can be found in a recent edited volume focused specifically on the in­ teractive roles of affect and cognition in intergroup perception (Mackie & Hamil­ ton, 1993). In conceptualizing how affect might influence intergroup perception, it is use­ ful to distinguish between different sources of affect in the intergroup situation. The affect experienced by the perceiver might have been generated prior to and independently of the intergroup context. Bodenhausen (1993) referred to this as incidental affect, and in this case we typically are concerned with the influence of the perceiver’s arousal or mood state on the formation and/or use of stereo­ types in processing information about group members. Alternatively, the affect experienced by the perceiver may have its origins in the intergroup context, be­ ing generated by one’s perceptions of or interaction with members of some out­ group. In this case, the affect is integral to the intergroup context (Bodenhausen, 1993). Here the perceiver’s preexisting stereotype, as well as other aspects of the intergroup situation, can be the catalyst for the experienced affect. In addi­ tion, the affect generated in this setting can influence subsequent information processing, which in turn can affect further the participant’s feelings, perceptions, and behavior. In the following sections, we discuss recent research on the effects of both incidental and integral affect in intergroup contexts. Incidental Affect and Stereotyping

One way that affect influences stereotyping is through the effect of the perceiver’s mood on the way information about groups and group members is processed. Research has shown that mood states can have both cognitive and motivational effects on social information processing (Forgas, 1992; Isen, 1984; Mackie, Asun­ cion, & Rosselli, 1992; Mackie & Worth, 1991; Schwarz, 1990). However, in any given situation, the motivational and cognitive consequences of mood might suggest different effects on information processing. From a motivational perspec­ tive, positive and negative moods have differing implications for information processing. Positive mood is predicted to decrease the thoroughness with which information is processed, either to prevent that information from distracting one away from a pleasant affective state or because the positive mood conveys the message that, because “all is well,” there is no need to focus on new material. Negative mood would stimulate more extensive processing of new information, either as a means of distracting oneself from the negative affect or because the negative mood informs the person that the current situation needs to be altered. From a cognitive perspective, mood influences processing by interfering with one’s ability to perform cognitive tasks. This interference may emanate from the mood state distracting one’s attention from the task at hand, or it may result from reduced cognitive capacity due to the mood consuming some cognitive resources.

44

HAMILTON AND SHERMAN

In either case, positive and negative moods should have much the same effect on these processes. Research by a number of investigators has demonstrated motivational and cog­ nitive effects of mood on both the formation and use of stereotypes. In these studies, subjects typically are induced into a positive, neutral, or negative mood state (e.g., by watching affectively toned videotaped material or by thinking about pleasant or unpleasant past experiences), and then, in a presumably separate ex­ periment, they are asked to process information pertinent to certain social groups. The effect of the induced mood on subjects’judgments, recall, processing times, and other relevant measures can then be determined. Some studies have examined the effects of mood on processes that would con­ tribute to the initial development of stereotypic responding. For example, Stroess­ ner and Mackie (1992, 1993) showed that subjects in positive and negative (compared with neutral) mood conditions were less able to detect the degree of variation among members of a group, and hence underestimated overall group variability. As noted in an earlier section, insensitivity to within-group differ­ ences enhances the likelihood that perceivers will generalize across group mem­ bers in attributing characteristics to the group (i.e., stereotyping). In the same vein, Wilder (1993; Wilder & Shapiro, 1989) showed that anxious subjects were less likely to differentiate among members of a group and assimilated a deviant member to the group as a whole. Similarly, research using an illusory correla­ tion paradigm (Stroessner et al. 1992; Hamilton, Stroessner, & Mackie, 1993) showed that both positive and negative mood states (compared with a neutral mood condition) reduced subjects’ sensitivity to the distinctiveness properties of the stimu­ lus information (i.e., the relative infrequency of the minority group performing undesirable behaviors), and hence these subjects did not form an illusory corre­ lation. In all of these cases, induced mood undermined the thoroughness with which group-descriptive information was processed. Interestingly, in one case, this effect created conditions increasing the likelihood of stereotyping (i.e., reduced detection of within-group variability), whereas in the other case, the effect decreased the likelihood of stereotype formation (i.e., through undermining the illusory correlation effect). Other studies investigated the effects of incidental affect on the use of existing stereotypes in perceptions of social groups (Bodenhausen, 1993; Forgas & Moylan, 1991; Hamilton et al., 1993; Kim & Baron, 1988; Mackie et al., 1989). One illustration is found in some recent work by Bodenhausen (1993). As dis­ cussed earlier, Bodenhausen (1988, 1990; Bodenhausen & Lichtenstein, 1987; Bodenhausen & Wyer, 1985) has argued that stereotypes can be used as a heuris­ tic or simplified processing strategy under conditions of limited processing ca­ pacity or reduced motivation. Drawing on research showing that affect has those effects on cognitive processing, he reasoned that induced mood might lead to in­ creased reliance on stereotypes in social judgments. In one study testing this hypothesis, Bodenhausen (1993) first induced happiness,

1.

STEREOTYPES

45

sadness, or anger and then had subjects read a description of a case in which one student physically assaulted another. For half of the subjects, the accused student’s name identified him as Hispanic, in which case the aggressive act was consistent with the stereotype of his ethnic group. The subjects’ task was to evaluate the student’s guilt. As predicted, subjects in happy and angry moods—affects hypothesized to reduce processing capacity and motivation—judged the case in stereotypic terms. That is, subjects in these conditions who read about a stereo­ typed defendant were more likely to find him guilty than were subjects reading about a student without a Hispanic surname. In contrast, this difference did not occur for subjects in a sad mood, consistent with Bodenhausen’s (1993) predic­ tion that sadness does not constrain processing and hence would not lead to reli­ ance on a group stereotype. (Although these results followed from Bodenhausen’s analysis, the issue of under what conditions sadness leads to more or less careful processing is a matter of debate; see Mackie & Worth, 1991; Stroessner et al., 1992; Stroessner & Mackie, 1992). As these studies illustrate, incidental affect can have consequences for both the nature and extent of the perceiver’s processing of group-relevant information. The ramifications of these effects for both stereotype formation and the use of existing stereotypes in social judgments will continue to be the focus of research activity. Integral Affect and Stereotyping

Incidental affect refers to prior feeling states that influence the perceiver’s per­ ceptions, because he or she brings them to the intergroup context. Therefore, incidental affect always precedes its effect on stereotyping, and the question con­ cerns how the affect influences cognitive processing in this context. The inter­ play between cognitive and affective factors involved in stereotyping becomes even more complex when the affect is generated in and is integral to the inter­ group context. In this situation, cognitive and affective factors each can have a causal effect on the other. That is, cognitive factors can initiate affective reac­ tions, and affective reactions in turn can influence the nature of cognitive process­ ing. We illustrate these points with some recent research examples. As discussed earlier, participants in many intergroup contexts react in terms of a distinction between in-group and out-group; we already discussed research relevant to several topics (e.g., in-group bias, out-group homogeneity effects) deriving from that distinction. Recently, Dovidio and Gaertner (1993) argued that in-group/out-group categorization automatically activates category-based affec­ tive reactions that directly influence perceptions of and expectations about inter­ actions with group members. For example, they have shown that subliminal priming with terms like we and they automatically generates positive and nega­ tive reactions that have evaluative consequences for otherwise neutral stimuli. These reactions also include differing expectations for how pleasant their inter­ actions with target persons would be.

46

HAMILTON AND SHERMAN

Affective consequences of intergroup perception extend beyond mere in­ group/out-group categorization. Whereas the stereotype literature typically has focused on the trait-like beliefs activated by perception of group memberships, several recent studies have shown that the stereotypes associated with various social groups can generate affective reactions as well (Dijker, 1987; Jackson & Sullivan, 1989; Vanman & Miller, 1993), For example, Dijker (1987) showed that his Dutch subjects had distinct patterns of emotional reactions (as measured by rating scales) to three ethnic out-groups living in the Netherlands. Moreover, Vanman and Miller (1993) demonstrated that psychophysiological measures could detect different patterns of facial muscular activity, reflecting positive and nega­ tive affective reactions, as they viewed members of different groups. Several studies found that these affective reactions are as (and sometimes more) effective predictors of intergroup attitudes as are stereotypic beliefs (Esses, Haddock, & Zanna, 1993; Stangor, Sullivan, & Ford, 1991; Stephan & Stephan, 1993). Other research has shown that intergroup interaction (real or anticipated) can generate affective responses that can influence the course of social interaction. Stephan and Stephan (1985) reviewed evidence supporting their hypothesis that intergroup contexts often generate anxiety that can lead to increased stereotyp­ ing, disrupted interaction, and avoidance of future interaction. Greater experience in intergroup contact may or may not alleviate these tendencies. Stephan and Ste­ phan (1985) found that increased intergroup contact was correlated negatively with these effects of intergroup anxiety, but Dijker (1987) found stronger (nega­ tive) emotional reactions to out-groups among those who had had more interac­ tion with them. Obviously, other factors can moderate these relationships. One relevant factor is the nature of the interaction context. For example, Wilder (1993; Wilder & Shapiro, 1989) found that competitive interaction generated anxietyrelated effects on intergroup perceptions, whereas noncompetitive interaction did not. Similarly, Fiske and Ruscher (1993) argued that the nature of the interde­ pendence relationship between group members can influence the extent to which out-group members are seen as disrupting the perceiver’s goal attainment in the interaction (see also Vanman & Miller, 1993). Can these effects of integral affect be overcome? Earlier, we discussed De­ vine’s (1989) hypothesis that cultural stereotypes of prominent ethnic groups are activated automatically when group members are encountered, and that lowprejudiced persons must counteract those reactions by applying their personal (nonprejudiced) beliefs about those groups. More recently, Devine and her col­ leagues (Devine & Monteith, 1993; Devine, Monteith, Zuwerink, & Elliot, 1991) have investigated the affective consequences of the discrepancy between people’s beliefs about how they would and how they should respond in intergroup con­ texts. Specifically, they argued that the low-prejudiced person’s failure to inhibit stereotypic responses to an out-group member activates a discrepancy between internalized nonprejudiced standards and actual responses. This discrepancy, which threatens the person’s nonprejudiced values and self-concept, produces negative

1.

STEREOTYPES

47

affective responses (e.g., guilt, self-criticism), which in turn can lead to efforts to control stereotypic responding. Through this process, persons can learn to respond in ways consistent with their nonprejudiced personal beliefs and standards. Most research on stereotype use has focused on processes that transpire in the stereotype holder and examine the cognitive and affective consequences of those beliefs. Much less research has considered the cognitive and affective dy­ namics of the target of stereotypic beliefs in social interaction. Recently, however, Crocker and Major (1989; Major & Crocker, 1993) have investigated the affec­ tive consequences of stereotypes for the way such target persons experience so­ cial interactions. Specifically, they showed that stereotyped target persons experience attributional ambiguity in interpreting the causes of others’ behavior toward them. For example, a negative evaluation of a minority person’s perfor­ mance received from a majority group member might be due to a valid assess­ ment of the target person’s poor performance. Alternatively, it could be attributed to the prejudicial attitudes of the evaluator. In the latter case, attributing the nega­ tive feedback to the evaluator’s prejudice buffers the affective impact of that feed­ back, and thereby protects one’s self-esteem. Similarly, a favorable evaluation might constitute legitimate praise for commendable performance or, alternative­ ly, it might reflect the evaluator’s effort to avoid appearing prejudiced. In this case, the positive effects of favorable evaluation would be muted, again diminishing the effect of feedback on one’s affective reactions and self-esteem. As a conse­ quence, both positive and negative feedback may have diminished affective im­ pact on minority group members. These authors reported a series of experiments illuminating the conditions that both augment and diminish this attributional am­ biguity and its effects on self-evaluations. The research reviewed here illustrates the variety of ways that affective and cognitive factors intermingle in determining social perceptions, emotional reac­ tions, and behavioral dynamics in intergroup contexts. This work effectively high­ lights the important and complex ways that stereotypes can influence, and be influenced by, affective concomitants in intergroup settings.

STEREOTYPE CHANGE

How can we change people’s stereotypes? Given the pervasiveness of stereotypes and the frequency with which they unfairly victimize members of ethnic groups, it becomes especially important to identify means by which stereotypic beliefs can be undermined or modified. Unfortunately, despite the importance of this question, the problem of changing stereotypes remains a very real and unsolved dilemma. Although the three conceptual orientations to stereotyping focus on different underlying mechanisms, they all share one common element relating to this is­ sue. In all three cases, the mechanisms that are emphasized to explain why stereo­

48

HAMILTON AND SHERMAN

types exist and persist are the reasons why stereotypes are difficult to change. We briefly illustrate this point for all three perspectives. In the psychodynamic approach, stereotypic beliefs about out-group members serve intraphysic needs, and hence are functionally important to the inner work­ ings of the perceiver’s mentality. For example, in classical psychodynamic the­ ory, stereotyping reflects the operation of defense mechanisms such as projection and displacement, mechanisms that are functional in keeping unacceptable sources of anxiety out of conscious awareness. Changing stereotypic beliefs might un­ dermine the effectiveness of these defensive processes, thereby creating new sources of anxiety. Similarly, for more contemporary motivational theories that emphasize the self-esteem benefits of viewing out-groups in negative terms, chang­ ing stereotypic beliefs about out-groups would remove a potentially important mechanism contributing to the person’s positive self-regard. In both cases, stereotyping serves an important function within the person’s overall personali­ ty. Consequently, simply presenting the person with evidence that some belief is wrong would hardly be effective. Thus, for these approaches, attempts to change stereotypes somehow must confront the question of how these needs are to be met and how these functions are to be served by alternate means. Otherwise, strategies toward change are likely to produce effects that are transitory at best. Similar difficulties confront efforts toward change from the sociocultural per­ spective. This viewpoint emphasizes the importance of the social environment (authority figures, peer groups, social roles, media portrayals, etc.) in the for­ mation and maintenance of stereotypic belief systems. These beliefs are acquired through social learning and are maintained through social reinforcement. Again, simply presenting the person with evidence that some belief is wrong hardly would be effective. Attempts to change the person’s beliefs somehow would require al­ tering elements of the person’s social environment, if not the social structure in which the person lives. Otherwise, the belief-maintaining supports would still be in place. For the cognitive approach, the story is much the same, although for different reasons . The importance of the categorization process in social perception seems fundamental, both for identifying the regularities and differences in one’s social environment and for simplifying the task of information processing in a complex stimulus world. Once a pattern of beliefs becomes associated with those categories, the resulting stereotypes often guide subsequent information processing. We have discussed numerous examples of how existing stereotypes can influence the way new information about groups and group members is processed. In virtually ev­ ery case, the resulting bias serves to maintain the status quo. The perceiver is more likely to attend to and notice stereotype-consistent information, to make stereotype-consistent inferences, to recall stereotype-consistent information, and so on. The overall consequence is that the perceiver “sees” a pattern of informa­ tion that seems to provide evidence for the “validity” of the beliefs that them­ selves influenced the way the information was processed. Once again, simply

1.

STEREOTYPES

49

presenting evidence that a particular belief is wrong seems unlikely to induce effective change. Thus, all three conceptual orientations effectively point to mechanisms that serve to maintain preexisting stereotypic beliefs, making change difficult. Yet, despite their persistence, we know that stereotypes do change, at least under cer­ tain conditions. In this regard, it is interesting that the same research that often is cited as documenting the persistence of stereotypes over time also presents evidence of dramatic change in stereotypes. This evidence comes from the well-known “Prince­ ton trilogy,” in which the same stereotype assessment procedures were used in three studies spanning some 35 years—Katz and Braly (1933) during the depres­ sion; Gilbert (1951) in the post-World War II years; and Karlins, Coffman, and Walters (1969) in the 1960s. In these studies, subjects were asked to indicate, on an adjective checklist, those attributes that were most characteristic of each of several national and ethnic groups. Attributes most commonly endorsed were taken to constitute the primary features of the stereotype of that group. One promi­ nent aspect of the findings of these studies is the remarkable stability of some of these stereotypes over a period of three and a half decades. For example, at each time period, the most frequently endorsed characteristics of Negroes included lazy, happy-go-lucky, and musical. However, this stability is only part of the story from this research. The most prominent features of the stereotype of Japanese changed from intelligent, industrious, and progressive in 1933 to imitative, sly, and extremely nationalistic in 1951, and then to ambitious, efficient, and loyal to family ties in 1967. The reason for these substantial shifts can be attributed, of course, to the intervention of World War II between the first and second as­ sessments, and to Japan’s subsequent reconstruction and realignment with the Unit­ ed States between the second and third assessments. Similar, although perhaps less dramatic, changes have occurred in stereotypes of African-Americans and women as a function of the civil rights and women’s movements, both of which gained momentum during the 1960s. These findings document that stereotypes do change under some conditions. The problem then becomes to identify those mechanisms that are important in causing and in mediating these changes. Models of Stereotype Change

Although our understanding of this issue is still far from adequate, social cogni­ tion research in the last decade has at least provided some conceptual bases for analyzing this process. These analyses focus on how information that disconfirms a stereotype might affect the preexisting stereotypic beliefs and when that effect is most likely to occur. Several models or mechanisms by which such change might take place have been proposed. The bookkeeping model (Rothbart, 1981) suggests a gradual course of change as increasing amounts of disconfirming in­ formation are encountered and accumulated. According to this view, each new

50

HAMILTON AND SHERMAN

piece of information—confirming or disconfirming—acquired about a target group leads to an adjustment in the stereotypic beliefs, either strengthening or weaken­ ing the existing stereotype by some modest amount. Hence, stereotypes are con­ stantly in a state of on-line revision as new information is incorporated. In this process, change would occur slowly as current beliefs are gradually adjusted in light of recent information and experience. A second process, which Rothbart (1981) called the conversion model, posits that change can occur suddenly and significantly in response to critically impor­ tant and convincing (to the perceiver) instances of stereotype disconfirmation. That is, some experience with out-group members, some challenging new piece of information about the group, or some group-relevant political or cultural event would be sufficiently compelling to lead the stereotype holder to substantially revise his or her beliefs about the target group. In contrast to the bookkeeping model’s slow, gradual change in response to accumulating bits of information, this process involves a rapid change of major magnitude (i.e., a conversion) in response to a dramatic disconfirmation experience. A third model of change, derived from cognitive research on category represen­ tations, focuses on a subtyping process (Brewer et al., 1981; Taylor, 1981). In this view, disconfirming information about members of a stereotyped group results in the differentiation of the large group (superordinate category) into subtypes (subordinate categories), with separate patterns of beliefs associated with each of the subtypes. In essence, the process is one of breaking down the large group into smaller subcategories. In contrast to the previous models, this process does not necessarily involve any actual change in preexisting beliefs, but instead posits a mechanism by which those initial beliefs would be applied less generally. Research Evidence

Patterns o f Disconfirming Information. The three models just described dis­ cuss alternate routes by which disconfirming information might bring about change in stereotypic beliefs. The situation addressed by all of them reflects important elements of everyday experiences. That is, we begin with a person who holds a stereotype about a particular target group. Then, over time and through vari­ ous experiences, the person encounters a number of individual members of that target group, some of whom behave in a manner consistent with stereotype-based expectancies, whereas the behavior of others seems to violate those expectan­ cies. In essence, the perceiver faces an information processing task in which the available information provides some mixture of confirmation and disconfirma­ tion of prior beliefs. Given this situation, what properties of that “mixture” are more or less likely to induce stereotype change, and why? This question has been the focus of some recent research. Several studies have used essentially the same paradigm to study this issue: Subjects learn information about several members of a stereotyped group, each

1.

STEREOTYPES

51

person being described by several items of information. Each information item either confirms or disconfirms the stereotype, or is unrelated to the stereotype. The manipulation of primary interest is whether the disconfirming items are (a) concentrated in describing one (or a small proportion) of the group members or (b) dispersed throughout the descriptions of all (or most) group members. The question of interest concerns how these patterns of disconfirming information af­ fect ratings of the group on stereotypic attributes. The results of these studies are not entirely consistent, and they have been interpreted by various authors as providing evidence in support of all three of the models discussed previously. For example, some studies (Gurwitz & Dodge, 1977; Hewstone, Johnston, & Aird, 1992) found less stereotypic ratings when the disconfirming information was concentrated in the descriptions of one or a few individuals. Because these individuals provide pronounced or dramatic vio­ lation of stereotypic expectancies, this evidence has been interpreted as support for the conversion model of stereotype change. On the other hand, other studies (Johnston & Hewstone, 1992; Weber & Crocker, 1983) obtained less evidence of stereotyping when the disconfirming items were dispersed across several group members than when concentrated in a few. The greater effectiveness of discon­ firming information under dispersed than concentrated conditions has been inter­ preted as support for the subtyping model. The argument here is that, when disconfirming items are concentrated in a few group members, those individuals will be grouped together in a subtype, and therefore their disconfirmation value will have relatively little effect on perceptions of the group as a whole. In fact, hierarchical clustering analyses indicate that disconfirmers are subgrouped more clearly when the disconfirming information is concentrated in a few members (Johnston & Hewstone, 1992). Research also indicates that these general patterns may be dependent on other factors, such as the number of group members presented to the subjects (Weber & Crocker, 1983, Experiments 1 & 4), the number of groups presented (Weber & Crocker, 1983, Experiment 4), and whether the stereotypic expectancy about the target group implies that group members are homogeneous or heterogeneous (Hewstone et al., 1992). It also remains unclear whether these patterns of dis­ confirming information have their primary effect on perceptions of stereotypeconsistent traits (Hewstone et al., 1992), stereotype-inconsistent traits (Johnston & Hewstone, 1992), or both (Weber & Crocker, 1983). Subtyping and Stereotype Change. The finding that stereotype-disconfirming information, particularly when concentrated in a few group members, can result in the creation of subtypes that are characterized by distinct sets of attributes demonstrates that people develop hierarchical systems of social categories. As these subtypes develop, stereotypes may form pertaining to them as well as to the superordinate category (Brewer et al., 1981; Devine & Baker, 1991). Given that disconfirming members can form a subtype, the question then arises: What

52

HAMILTON AND SHERMAN

is the relationship between subtyping and stereotype change? That is, how does the formation of subtypes affect the stereotype of the overall group? Subtyping commonly is cited as one of the prominent models of stereotype change (Hewstone, 1989; Hewstone et al., 1992; Johnston & Hewstone, 1992; Weber & Crocker, 1983). What is the nature of that change? If stereotyping rests on categorization, as argued earlier, and if category-based beliefs are applied to individuals as they are categorized, then the use of subtypes certainly alters the nature of the stereotyping that occurs. That is, if an individual is viewed as a member of a subtype, rather than being categorized as a member of the superordinate group, then this can have meaningful consequences. First, the beliefs associated with several of the subtypes may differ dramatically from those ascribed to the superordinate group (Brewer et al., 1981; Devine & Baker, 1991). If so, then the attributes assumed to characterize that individual could be substantially different from those inferred from the overall group stereotype, as well as from alternative subtypes. Second, although the use of subtypes still would constitute stereotyping (i.e., ascribing attributes on the basis of group member­ ship), those attributes would be ascribed to a smaller, narrower range of persons (e.g., career women, urban Blacks, southern Democrats), rather than to the su­ perordinate category as a whole (women, Blacks, Democrats). Thus, the over­ generalization that characterizes stereotyping could be reduced significantly (Hamilton, 1981). These are important consequences that derive from the formation and use of subtypes. Nevertheless, there still will be occasions when the superordinate category will be considered an adequate basis for categorization and hence will be employed. The question then remains, does the presence of subtypes affect the nature of the beliefs ascribed to the overall group? Although this question has rarely been discussed in the stereotype literature, several possible answers can be gleaned from various discussions of subtyping. One possibility is that the subtypes become sufficiently strong that they dilute the beliefs held about the superordinate category as a whole. For example, the Brewer et al. (1981) research on stereotypes of the elderly provided evidence that subjects held three different stereotypes about different subgroups of elderly persons. Each of these conceptions was richly articulated and widely shared in the subject population, yet the three were clearly different from each other. Moreover, these subjects had very little consensus about the attributes shared by old people in general (i.e., a stereotype that generalizes across the subtypes). Thus, the presence of well-developed (and presumably, frequently used) subtypes apparently has diluted the content of the stereotype of the generic category. To the extent that this happens, subtyping indeed would lead to stereotype change. However, in other instances, this consequence is less evident. For example, Devine and Baker (1991) attempted to define and measure several subtypes of Blacks, including ghetto Blacks, Black athletes, streetwise Blacks, welfare Blacks, and businessman Blacks. They found that Black athletes and Black businessmen

1.

STEREOTYPES

53

formed distinct subtypes about whom subjects held clear, but differential, be­ liefs. On the other hand, ghetto Blacks, streetwise Blacks, and welfare Blacks were less distinct subgroups and shared overlapping attributes among themselves and with the generic stereotype of Blacks. In this case, it appears that certain subgroups (athletes, businessmen) are viewed as meaningful and unique subgroups of the overall category, whereas the stereotype of the superordinate category still applies, in large part, to most other group members (with other subgroups reflect­ ing minor differences). Finally, there may be cases in which highly specific, fairly narrow subtypes develop to represent disconfirming exceptions (Pettigrew, 1981) for the specific purpose of preserving the generic stereotype (Weber & Crocker, 1983). That is, if disconfirming cases can be separated into a distinct category that isolates them from the predominant group, their atypicality might diminish their perceived relevance for perceptions of the group as a whole. Thus, rather than producing stereotype change, the disconfirming instances become their own group, and there­ by pose no challenge to the preexisting belief structure. In this case, subtyping becomes a mechanism contributing to stereotype preservation, rather than stereo­ type change. The implications of such “re-fencing” (Allport, 1954) are developed further in the next subsection. Nature of Disconfirming Instances. The pattern in which disconfirming in­ formation is distributed among group members is only one of several properties of information that can affect the extent or likelihood of stereotype change. Another important factor concerns the extent to which a stereotype-disconfirming individual is perceived as being a “good” member of the social category about which the stereotype is held. As Rothbart and John (1985) pointed out, an individual per­ son can be categorized in terms of any number of groups. They argued that if a person strongly violates the perceiver’s conception of a particular group to which the target person belongs, the person will not be viewed as a member of that category and instead will be categorized in terms of some alternative group membership. For example, the football player who earns straight As, prefers fine wine, classical music, and English literature to beer, hard rock, and comic books, and shows warmth and sensitivity in his interpersonal relationships will not be categorized as a “jock,” but rather as a “sophisticate” who happens to play a sport. That is, because he so completely violates the “jock” stereotype, he is not per­ ceived as a member of that category. Consequently, the stereotype-disconfirming information that he provides is unlikely to alter the perceiver’s stereotype of jocks; because he is not categorized as a jock, this information is not viewed as relevant to that stereotype. (Instead, his interest in playing football may produce some loosening of the perceiver’s stereotype of “sophisticates,” the social category that has been invoked by this individual.) Thus, there is a delicate balance between the extent to which an individual provides stereotype-disconfirming information and the ability of that information to have impact on the preexisting stereotype.

54

HAMILTON AND SHERMAN

Rothbart and John (1985) argued that, to be an effective implement of change, the individual must (a) provide some stereotype-disconfirming information but (b) be perceived as a “good” member of the target group. Experimental evidence supporting this argument was offered by Rothbart and Lewis (1988). They presented subjects with a number of exemplars of four geo­ metric shapes (rectangles, triangles, etc.). For each shape, half of the exemplars were “poor” instances of the category (e.g., a rectangle where height is consider­ ably longer than width), whereas the other half were “good” instances of the category (e.g., an equilateral triangle). For each shape, the “good” and “poor” exemplars were of different color, so that color was correlated perfectly with goodness of fit. Subsequently, subjects were asked to estimate how many instances of each shape they had seen in each color. Subjects estimated that they had seen significantly more of the “good” than of the “poor” examples of each category. These findings have important implications for efforts to change stereotypes. Presenting disconfirming information will not be enough. In fact, presenting ex­ amples of individuals who strongly disconfirm stereotypic expectancies can be expected to have little, if any, effect, because those individuals will not be viewed as category members. Instead, group members who are in many respects represent­ ative of the target group, but who provide some disconfirmation of the stereo­ type, may be more effective in inducing change in stereotypic beliefs. Nature of Stereotypic Attributes. Another factor that can influence the likeli­ hood of stereotype change is the nature of the attributes comprising the stereo­ type (Hewstone, 1989). As Rothbart and Park (1986) showed, traits vary considerably in their susceptibility to change. Trait concepts are abstractions based on behavioral manifestations. For some traits, these behavioral manifestations are clearly evident and frequently observable, whereas for other traits those manifestations occur less frequently and are less apparent. For example, a per­ son’s cleanliness or dirtiness is readily observable and is apparent whenever that person is encountered; therefore, it is fairly easy for the perceiver to assess the validity of such a characterization and, if appropriate, to change one’s belief about the person. On the other hand, for many other traits, such opportunities for vali­ dation are less frequent and are more difficult, rendering change in such a belief less likely. For example, consider the difficulty of disconfirming a belief that a person is devious. Behavioral manifestations that would provide clear evidence contradictory to this belief are difficult to imagine, would occur only under un­ usual (infrequent) circumstances, and often would be open to alternative inter­ pretations that might undermine their potential for disconfirmation. In fact, even the absence of confirmatory manifestations can be viewed as substantiating the belief: the less one sees evidence of the person’s deviousness, the more effec­ tive he or she appears to be at it. To the extent that a group’s stereotype is composed of such traits, the more difficult it will be to induce change in those beliefs.

1.

STEREOTYPES

55

Similar considerations derive from Reeder and Brewer’s (1979) analyses of the asymmetries in inferences from behavior to trait. For example, intelligent people can do stupid things, as well as smart things, but if a person is stupid, he or she does not have the capacity to do highly intelligent things. Similarly, dishonest people can and do engage in honest, as well as dishonest, behaviors. However, the honest person does not have such flexibility; as soon as he or she engages in a dishonest act, he or she is no longer an honest person. These asym­ metries have implications for the likelihood of stereotype change. For instance, if a stereotype characterizes a group as dishonest, then honest behaviors by group members will not necessarily be viewed as compelling evidence against that be­ lief, because dishonest persons are not always dishonest and, in fact, often en­ gage in honest behaviors. Finally, traits also differ in their breadth or narrowness (Hampson, John, & Goldberg, 1986). For example, the trait “responsible” is a broad trait that en­ compasses many diverse behaviors that might manifest this quality. In contrast, the trait “punctual” also refers to behavior that is responsible, but it represents a narrower, more restrictive, and more constrained domain of behavioral manifestations. The effects of these differences for stereotyping were demonstrated by Hamilton, Gibbons, Stroessner, and Sherman (1992), who analyzed subjects’ ratings of liked and disliked nationalities on broad and narrow traits that differed in desirability. Not surprisingly, liked groups were rated highly on desirable traits and disliked groups were rated highly on undesirable traits. More interestingly, when subjects rated the groups on traits that were incompatible with their overall evaluations of the groups —rating liked groups on undesirable traits, or disliked groups on desirable traits—they made higher ratings on narrow than on broad attributes. That is, subjects acknowledged that a liked nationality possessed some undesirable attributes, and that disliked groups did have some favorable quali­ ties, but these exceptions were confined to narrow traits that would pose less challenge to the preexisting group evaluation. Although the research of Rothbart and Park (1986), Reeder (1985), and Hamil­ ton et al. (1992) was not focused on issues of stereotype change directly, it points to important properties of trait-based beliefs that would make them more or less amenable to change. Given that group stereotypes are, to a considerable extent, composed of such trait-based beliefs, these results point to important considera­ tions for an analysis of stereotype change, as well as indicating important avenues of future research on this topic. Outcome Bias and Stereotype Change. It is obvious that stereotypes are difficult to change, and, in discussing this literature, we have encountered numer­ ous reasons for their resistance to change even in the face of counterstereotypic information. From the cognitive perspective, the mechanisms that promote the use of stereotypes can undermine the effectiveness of efforts to change stereo­ typic beliefs. These mechanisms include processes and biases that, in many

56

HAMILTON AND SHERMAN

circumstances, are functional and adaptive, even if they reflect less than optimal information processing. By and large, the cognitive processes we discussed serve to maintain the beliefs that generated, or biased, that processing from the outset. Recently, however, Mackie and her colleagues (Mackie, Allison, Worth, & Asun­ cion, 1992a, 1992b; Mackie, Worth, & Allison, 1990) have pursued a research program exploring the ways in which biases in information processing might be used to promote the perception that groups have changed over time, thereby facilitating the modification of stereotypic beliefs. One of the best documented findings in the social perception literature is the correspondence bias—the tendency to infer that persons possess disposi­ tions that correspond to the manifest properties of their behaviors, even when this behavior is clearly constrained by situational factors (Jones, 1991). This phenomenon also has been demonstrated for group perceptions (Allison & Mes­ sick, 1985). The Mackie et al. research capitalized on the fact that a perceiv­ er’s correspondent inferences can be biased by the outcome of the observed performance. For example, Mackie et al. (1990) showed that when a group of college students was described as having qualified for a “college bowl” com­ petition, they were rated as being more intelligent than when they failed to qualify —a simple correspondent inference. However, this effect occurred (a) even when the group’s success or failure was determined by arbitrary changes in the decision rules governing the competition, and (b) even when the group’s actual behavioral performance was identical in the success and failure condi­ tions. More recently, this research has been extended in meaningful ways. First, Mackie et al. (1992b) showed that this outcome bias can generate biased infer­ ences about groups even when those inferences are counter-stereotypic. Specifi­ cally, these effects occurred even when Asian-Americans were said to fail and Blacks were said to succeed in an intelligence-related competition (outcomes that are counter to the stereotypes held in the subject population). Finally, Mackie et al. (1992a) showed that outcome-biased inferences based on the performance of group members can generalize to members of the stereotyped group as a whole. Specifically, subjects who had made outcome-biased inferences about the intelli­ gence of a group of eight Black or Asian-American college students were later asked to assess the intelligence of Whites, Blacks, and Asians in general. The results showed a modest, but significant, generalization of the outcome-biased inferences to the target group as a whole. The research by Mackie and her colleagues demonstrated that cognitive biases in the way group-descriptive information is processed (in this case, in­ ferences biased by performance outcomes) can influence judgments of the group as a whole, including the perception of change. These findings raise the possibil­ ity that other biasing factors also may be utilized as a means of inducing change in perceptions of, and ultimately beliefs about, stereotyped groups.

1.

STEREOTYPES

57

A FINAL COMMENT

In preparing this chapter, we have surveyed a sizeable literature addressing a variety of topics related to the nature and use of stereotypes. In doing so, we were impressed by a number of features of this work. First, the pure magnitude of this literature is impressive, particularly when one realizes that a considerable proportion of the research we discussed was reported within the last 15 years. Clearly, this has been a vigorous area of research, and clearly the study of stereo­ types is one domain that has been influenced in important ways by developments in social cognition. Second, we were impressed by the diversity of topics rele­ vant to stereotyping that have been investigated during this period. This is not a narrow, singularly focused research area. We believe our review reveals the multifaceted nature of this topic and of how it is being explored empirically. Third, we were impressed by the sophistication of this research area, both conceptually and empirically. Theoretical models have been developed, debated, and tested that are rather specific about the nature of cognitive representations or about the processing of different kinds of information. Also the experimental strategies and research technology used to test those theoretical ideas have evolved at a rapid pace. Fourth, we were impressed by the broadening of this conceptual approach to stereotyping. In contrast to a decade ago, it is no longer a purely cognitive analysis, but instead seeks to understand the role of affective and motivational processes as well. In all of these fundamental respects, research on stereotypes has made impressive progress. Social scientists have been studying the nature and functioning of stereotypes for several decades. They have approached this topic from several conceptual orientations and have used a variety of research tools in conducting their inquiry. It is clear that we are not very far down the path toward achieving a full under­ standing of this topic. Nevertheless, the breadth, depth, and vitality of the research we have reviewed provide encouraging omens of continued progress. As it has in the past, the specific nature of theorizing and research will continue to change and evolve. The research we have reviewed provides a solid foundation on which future developments can build.

ACKNOWLEDGMENTS

Preparation of this chapter was supported by NIMH Grant MH 40058 to the first author. The second author was supported by a National Science Foundation graduate fellowship. The authors thank Diane M. Mackie, Steven J. Sherman, and Robert S. Wyer, Jr., for their comments on preliminary versions of this chapter.

58

HAMILTON AND SHERMAN

REFERENCES Allison, S. T., & Messick, D. M. (1985). The group attribution error. Journal o f Experimental So­ cial Psychology, 21, 563-579. Allport, G. W. (1954). The nature o f prejudice. Cambridge, MA: Addison-Wesley. Andersen, S. M ., & Cole, S. W. (1990). “Do I know you?” The role of significant others in general social perception. Journal o f Personality and Social Psychology, 59, 384-399. Andersen, S. M ., & Klatzky, R. L. (1987). Traits and social stereotypes: Levels of categorization in person perception. Journal o f Personality and Social Psychology, 53, 235-246. Ashmore, R. D. (1981). Sex stereotypes and implicit personality theory. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 37-81). Hillsdale, NJ: Lawrence Erlbaum Associates. Ashmore, R. D., & Del Boca, F. K. (1979). Sex stereotypes and implicit personality theory: Toward a cognitive-social psychological conception. Sex Roles, 5, 219-248. Ashmore, R. D., & Del Boca, F. K. (1981). Conceptual approaches to stereotypes and stereotyping. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 1-35). Hillsdale, NJ: Lawrence Erlbaum Associates. Banaji, M. R., & Greenwald, A. G. (in press). Implicit stereotyping and prejudice. In M. P. Zanna & J. M. Olson (Eds.), The psychology o f prejudice, the Ontario Symposium (Vol. 7). Hillsdale, NJ: Lawrence Erlbaum Associates. Bargh, J. A., Bond, R. N., Lombardi, W. L., & Tota, M. E. (1986). The additive nature of chronic and temporary sources of construct accessibility. Journal o f Personality and Social Psychology, 50, 869-879. Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and social perception: The influence of trait information presented outside of conscious awareness on impression forma­ tion. Journal o f Personality and Social Psychology, 43, 437-449. Barsalou, L. W. (1990). On the indistinguishability of exemplar memory and abstraction in category representation. In T. K. Srull & R. S. Wyer, Jr. (Eds.), Advances in social cognition (Vol. 3, pp. 61-88). Hillsdale, NJ: Lawrence Erlbaum Associates. Bechtold, A., Naccarato, M. E ., & Zanna, M. P. (1986). Need fo r structure and the prejudicediscrimination link. Paper presented at the annual meeting of the Canadian Psychological Associa­ tion, Toronto, Ontario. Biernat, M ., Manis, M ., & Nelson, T. F. (1991). Stereotypes and standards of judgment. Journal o f Personality and Social Psychology, 60, 485-499. Billig, M ., & Tajfel, H. (1973). Social categorization and similarity in intergroup behavior. Europe­ an Journal o f Social Psychology, 3, 27-52. Bodenhausen, G. V. (1988). Stereotypic biases in social decision making and memory: Testing process models of stereotype use. Journal o f Personality and Social Psychology, 55, 726-737. Bodenhausen, G. V. (1990). Stereotypes as judgmental heuristics: Evidence of circadian variations in discrimination. Psychological Science, 1, 319-322. Bodenhausen, G. V. (1993). Emotion, arousal, and stereotypic judgments: A heuristic model of af­ fect and stereotyping. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 13-37). San Diego, CA: Academic Press. Bodenhausen, G. V., & Lichtenstein, M. (1987). Social stereotypes and information processing strate­ gies: The impact of task complexity. Journal o f Personality and Social Psychology, 52, 871-880. Bodenhausen, G. V., & Wyer, R. S., Jr. (1985). Effects of stereotypes on decision making and in­ formation processing strategies. Journal o f Personality and Social Psychology, 48, 267-282. Bower, G. H., & Karlin, M. B. (1974). Depth of processing pictures of faces and recognition memory. Journal o f Experimental Psychology, 4, 751-757. Brewer, M. B. (1979). In-group bias in the minimal intergroup situation: A cognitive-motivational analysis. Psychological Bulletin, 86, 307-324.

1.

STEREOTYPES

59

Brewer, M. B. (1988). A dual process model of impression formation. In T. K. Srull & R. S. Wyer, Jr. (Eds.), Advances in social cognition (Vol. 1, pp. 1-36). Hillsdale, NJ: Lawrence Erlbaum As­ sociates. Brewer, M. B. (1993). Social identity, distinctiveness, and in-group homogeneity. Social Cognition, 77, 150-164. Brewer, M. B., Dull, V., & Lui, L. (1981). Perceptions of the elderly: Stereotypes and prototypes. Journal o f Personality and Social Psychology, 41, 656-670. Brigham, J. C. (1971). Ethnic stereotypes. Psychological Bulletin, 76, 15-33. Brooks, L. R. (1978). Non-analytic concept formation and memory for instances. In E. Rosch & B. Lloyd (Eds.), Cognition and categorization (pp. 169-211). Hillsdale, NJ: Lawrence Erlbaum Associates. Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64, 123-152. Campbell, D. T. (1965). Ethnocentric and other altruistic motives. In D. Levine (Ed.), Nebraska symposium on motivation (pp. 283-311). Lincoln, NE: University of Nebraska Press. Cantor, N., & Mischel, W. (1979). Prototypes in person perception. In L. Berkowitz (Ed.), Ad­ vances in experimental social psychology (Vol. 12, pp. 3-52). New York: Academic Press. Carlston, D. E. (1980). The recall and use of traits and events in social inference processes. Journal o f Experimental Social Psychology, 16, 303-328. Chapman, L. J. (1967). Illusory correlation in observational report. Journal o f Verbal Learning and Verbal Behavior, 6, 151-155. Clary, E. G., & Tesser, A. (1983). Reactions to unexpected events: The naive scientist and interpre­ tive activity. Personality and Social Psychology Bulletin, 9, 609-620. Cohen, C. E. (1981). Person categories and social perception: Testing some boundaries of the processing effects of prior knowledge. Journal o f Personality and Social Psychology, 40, 441-452. Crocker, J., & Major, B. (1989). Social stigma and self-esteem: The self-protective properties of stigma. Psychological Review, 96, 608-630. Darley, J. M ., & Gross, P. H. (1983). A hypothesis-confirming bias in labeling effects. Journal o f Personality and Social Psychology, 44, 20-33. Deaux, K., Winton, W ., Crowley, M ., & Lewis, L. L. (1985). Level of categorization and content of gender stereotypes. Social Cognition, 3, 145-167. Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Jour­ nal o f Personality and Social Psychology, 56, 5-18. Devine, P. G., & Baker, S. M. (1991). Measurement of racial stereotype subtyping. Personality and Social Psychology Bulletin, 17, 44-50. Devine, P. G., & Monteith, M. J. (1993). The role of discrepancy-associated affect in prejudice reduction. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Inter­ active processes in group perception (pp. 317-344). San Diego, CA: Academic Press. Devine, P. G., Monteith, M. J., Zuwerink, J. R., & Elliot, A. J. (1991). Prejudice with and without compunction. Journal o f Personality and Social Psychology, 60, 817-830. Diehl, M. (1990). The minimal group paradigm: Theoretical explanations and empirical findings. In W. Stroebe & M. Hewstone (Eds.), European review o f social psychology (Vol. 1, pp. 263-292). Chichester, England: Wiley. Dijker, A. G. M. (1987). Emotional reactions to ethnic minorities. European Journal o f Social P sy­ chology, 17, 305-326. Dovidio, J. F., Evans, N., & Tyler, R. B. (1986). Racial stereotypes: The contents of their cognitive representation. Journal o f Experimental Social Psychology, 22, 22-37. Dovidio, J. F., & Gaertner, S. L. (1993). Stereotypes and evaluative intergroup bias. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group p e r­ ception (pp. 167-193). San Diego, CA: Academic Press. Duncan, B. L. (1976). Differential social perception and attribution of intergroup violence: Testing the lower limits of stereotyping of blacks. Journal o f Personality and Social Psychology, 34, 590-598.

60

HAMILTON AND SHERMAN

Esses, V. M ., Haddock, G., & Zanna, M. P. (1993). Values, stereotypes, and emotions as deter­ minants of intergroup attitudes. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 137-166). San Diego, CA: Aca­ demic Press. Fazio, R. H., Lenn, T. M ., & Effrein, E. A. (1984). Spontaneous attitude formation. Social Cogni­ tion, 2, 217-234. Fiedler, K. F. (1991). The tricky nature of skewed frequency tables: An information loss account of distinctiveness-based illusory correlations. Journal o f Personality and Social Psychology, 60, 24-36. Fiske, S. T. (1982). Schema-triggered affect: Applications to social perception. In M. S. Clark & S. T. Fiske (Eds.), Affect and cognition (pp. 55-78). Hillsdale, NJ: Lawrence Erlbaum As­ sociates. Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression formation, from category-based to individuating processes: Influences of information and motivation on attention and interpreta­ tion. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 1-74). New York: Academic Press. Fiske, S. T., Neuberg, S. L., Beattie, A. E., & Milberg, S. J. (1987). Category-based and attributebased reactions to others: Some informational conditions of stereotyping and individuating processes. Journal o f Experimental Social Psychology, 23, 399-427. Fiske, S. T., & Pavelchak, M. A. (1986). Category-based versus piecemeal-based affective responses: Developments in schema-triggered affect. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook o f motivation and cognition: Foundations o f social behavior (Vol. 1, pp. 167-203). New York: Guilford. Fiske, S. T., & Ruscher, J. B. (1993). Negative interdependence and prejudice: Whence the affect? In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 239-268). San Diego, CA: Academic Press. Fiske, S. T., & Taylor, S. E. (1991). Social cognition. New York: McGraw-Hill. Forgas, J. P. (1992). Affect and social perception: Research evidence and an integrative theory. In W. Stroebe & M. Hewstone (Eds.), European review o f social psychology (Vol. 3, pp. 183-223). Chichester, England: Wiley. Forgas, J. P., & Moylan, S. J. (1991). Affective biases in stereotype judgments. Cognition and Emotion, 5, 379-395. Fried, L. S., & Holyoak, K. J. (1984). Induction of category distributions: A framework for classifi­ cation learning. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 10, 234-257. Gilbert, D. T., & Hixon, J. G. (1991). The trouble of thinking: Activation and application of stereo­ typic beliefs. Journal o f Personality and Social Psychology, 60, 509-517. Gilbert, G. M. (1951). Stereotype persistence and change among college students. Journal o f Abnor­ mal and Social Psychology, 46, 245-254. Gilovich, T. (1981). Seeing the past in the present: The effect of associations to familiar events on judgments and decisions. Journal o f Personality and Social Psychology, 40, 797-808. Gurwitz, S. B., & Dodge, K. A. (1977). Effects of confirmations and disconfirmations on stereotypebased attributions. Journal o f Personality and Social Psychology, 35, 495-500. Hamilton, D. L. (1979). A cognitive-attributional analysis of stereotyping. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 12, pp. 53-84). New York: Academic Press. Hamilton, D. L. (1981). Stereotyping and intergroup behavior: Some thoughts on the cognitive ap­ proach. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and inter group behavior (pp. 333-353). Hillsdale, NJ: Lawrence Erlbaum Associates. Hamilton, D. L., Driscoll, D. M ., & Worth, L. T. (1989). Cognitive organization of impressions: Effects of incongruency in complex representations. Journal o f Personality and Social Psycholo­ gy, 57, 925-939.

1.

STEREOTYPES

61

Hamilton, D. L., Dugan, P. M., & Trolier, T. K. (1985). The formation of stereotypic beliefs: Fur­ ther evidence for distinctiveness-based illusory correlations. Journal o f Personality and Social Psy­ chology, 48, 5-17. Hamilton, D. L., Gibbons, P., Stroessner, S. J., & Sherman, J. W. (1992). Stereotypes and lan­ guage use. In K. Fiedler & G. R. Semin (Eds.), Language and social cognition (pp. 102-128). Newbury Park, CA: Sage. Hamilton, D. L., & Gifford, R. K. (1976). Illusory correlation in interpersonal perception: A cogni­ tive basis of stereotypic judgments. Journal o f Experimental Social Psychology, 12, 392-407. Hamilton, D. L., & Mackie, D. M. (1990). Specificity and generality in the nature and use of stereo­ types. In T. K. Srull & R. S. Wyer, Jr. (Eds.), Advances in social cognition (Vol. 3, pp. 99-110). Hillsdale, NJ: Lawrence Erlbaum Associates. Hamilton, D. L., & Rose, T. L. (1980). Illusory correlation and the maintenance of stereotypic be­ liefs. Journal o f Personality and Social Psychology, 39, 832-845. Hamilton, D. L., & Sherman, S. J. (1989). Illusory correlations: Implications for stereotype theory and research. In D. Bar-Tal, C. F. Graumann, A. W. Kruglanski, & W. Stroebe (Eds.), Stereo­ types and prejudice: Changing conceptions (pp. 59-82). New York: Springer-Verlag. Hamilton, D. L., Sherman, S. J., & Ruvolo, C. M. (1990). Stereotype-based expectancies: Effects on information processing and social behavior. Journal o f Social Issues, 46(2), 35-60. Hamilton, D. L., Stroessner, S. J., & Driscoll, D. M. (in press). Social cognition and the study of stereotypes. In P. G. Devine, D. L. Hamilton, & T. M. Ostrom (Eds.), Social cognition: Con­ tributions to classic issues in social psychology. New York: Springer-Verlag. Hamilton, D. L., Stroessner, S. J., & Mackie, D. M. (1993). The influence of affect on stereotyp­ ing: The case of illusory correlations. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cogni­ tion, and stereotyping: Interactive processes in group perception (pp. 39-61). San Diego, CA: Academic Press. Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview of the cogni­ tive approach. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 127-163). Orlando, FL: Academic Press. Hampson, S. E., John, O. P., & Goldberg, L. R. (1986). Category breadth and hierarchical struc­ ture in personality: Studies in asymmetries in judgments of trait implications. Journal o f Personal­ ity and Social Psychology, 51, 37-54. Hastie, R. (1984). Causes and effects of causal attribution. Journal o f Personality and Social Psy­ chology, 46, 44-56. Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the judgment task is memory-based or on-line. Psychological Review, 93, 258-268. Hayes-Roth, B., & Hayes-Roth, F. (1977). Concept learning and the recognition and classification of exemplars. Journal o f Verbal Learning and Verbal Behavior, 16, 321-338. Heilman, M. E. (1984). Information as a deterrent against sex discrimination: The effects of appli­ cant sex and information type on preliminary employment decisions. Organizational Behavior and Human Performance, 33, 174-186. Hewstone, M. (1989). Changing stereotypes with disconfirming information. In D. Bar-Tal, C. F. Graumann, A. W. Kruglanski, & W. Stroebe (Eds.), Stereotypes and prejudice: Changing con­ ceptions (pp. 207-223). New York: Springer-Verlag. Hewstone, M., Johnston, L., & Aird, P. (1992). Cognitive models of stereotype change (2): Percep­ tions of homogeneous and heterogeneous groups. European Journal o f Social Psychology, 22, 235-249. Higgins, E. T ., Bargh, J. A., & Lombardi, W. (1985). The nature of priming effects on categoriza­ tion. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 11, 59-69. Higgins, E. T., & King, G. A. (1981). Accessibility of social constructs: Information-processing consequences of individual and contextual variability. In N. Cantor & J. F. Kihlstrom (Eds.), Per­ sonality, cognition, and social interaction (pp. 69-122). Hillsdale, NJ: Lawrence Erlbaum Associates.

62

HAMILTON AND SHERMAN

Higgins, E. T., Rholes, W. S., & Jones, C. R. (1977). Category accessibility and impression forma­ tion. Journal o f Experimental Social Psychology, 13, 141-154. Hintzman, D. L. (1986). “Schema abstraction” in a multiple-trace memory model. Psychological Review, 93, 411-428. Homa, D., Dunbar, S., & Nohre, L. (1991). Instance frequency, categorization and the modulating effect of experience. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 17, 444-458. Isen, A. M. (1984). Toward understanding the role of affect in cognition. In R. S. Wyer, Jr., & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 3, pp. 179-236). Hillsdale, NJ: Lawrence Erlbaum Associates. Jackson, L. A., & Sullivan, L. A. (1989). Cognition and affect in evaluations of stereotyped group members. The Journal o f Social Psychology, 129, 659-672. Jacoby, L. L., Kelley, C. M., Brown, J., & Jasechko, J. (1989). Becoming famous overnight: Limits on the ability to avoid unconscious influences of the past. Journal o f Personality and Social P sy­ chology, 56, 326-338. Johnson, C., & Mullen, B. (in press). Evidence for the accessibility of paired distinctiveness in distinctiveness-based illusory correlation in stereotyping. Personality and Social Psychology Bulletin. Johnston, L., & Hewstone, M. (1992). Cognitive models of stereotype change (3): Subtyping and the perceived typicality of disconfirming group members. Journal o f Experimental Social Psychology, 28, 360-386. Jones, E. E. (1991). Interpersonal perception. New York: W. H. Freeman. Jones, E. E., Wood, G. C., & Quattrone, G. A. (1981). Perceived variability of personal charac­ teristics in in-groups and out-groups: The role of knowledge and evaluation. Personality and So­ cial Psychology Bulletin, 7, 523-528. Jones, R. A., & Ashmore, R. D. (1973). The structure of intergroup perception: Categories and dimensions in views of ethnic groups and adjectives used in stereotype research. Journal o f Per­ sonality and Social Psychology, 25, 428-438. Judd, C. M ., & Park, B. (1988). Out-group homogeneity: Judgments of variability at the individual and group levels. Journal o f Personality and Social Psychology, 54, 778-788. Kahneman, D., & Miller, D. T. (1986). Norm theory: Comparing reality to its alternatives. Psycho­ logical Review, 93, 136-153. Karasawa, M ., & Brewer, M. B. (1992). Category size and judgm ents o f variability: The effects o f seeing the trees in the forest. Unpublished manuscript, University of California, Los Angeles. Karlins, M ., Coffman, T. L., & Walters, G. (1969). On the fading of social stereotypes: Studies in three generations of college students. Journal o f Personality and Social Psychology, 13, 1-16. Kashima, E. S., & Kashima, Y. (1993). Perceptions of general variability of social groups. Social Cognition, 11, 1-21. Katz, D., & Braly, K. (1933). Racial stereotypes in one hundred college students. Journal o f Abnor­ m al and Social Psychology, 28, 280-290. Katz, I. (1981). Stigma: A social psychological analysis. Hillsdale, NJ: Lawrence Erlbaum Associates. Katz, P. A. (1976). The acquisition of racial attitudes in children. In P. A. Katz (Ed.), Towards the elimination o f racism (pp. 125-154). New York: Pergamon. Kim, H-S., & Baron, R. S. (1988). Exercise and the illusory correlation: Does arousal heighten stereo­ typic processing? Journal o f Experimental Social Psychology, 24, 366-380. Klein, S. B., Loftus, J., Trafton, J. G., & Fuhrman, R. W. (1992). Use of exemplars and abstrac­ tions in trait judgments: A model of trait knowledge about the self and others. Journal o f Personal­ ity and Social Psychology, 63, 739-753. Kraus, S., Ryan, C. S., Judd, C. M ., Hastie, R., & Park, B. (1993). Use of mental frequency distri­ butions to represent variability among members of social categories. Social Cognition, 11, 22-43. Krueger, J., & Rothbart, M. (1988). Use of categorical and individuating information in making inferences about personality. Journal o f Personality and Social Psychology, 55, 187-195.

1.

STEREOTYPES

63

Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480-498. Kunda, Z., & Sherman-Williams, B. (1992). Stereotypes and the construal of individuating informa­ tion. Personality and Social Psychology Bulletin, 19, 90-99. Lewicki, P. (1985). Nonconscious biasing effects of single instances of subsequent judgments. Jour­ nal o f Personality and Social Psychology, 48, 563-574. Lingle, J. H., & Ostrom, T. M. (1979). Retrieval selectivity in memory-based impression judgments. Journal o f Personality and Social Psychology, 37, 180-194. Linville, P. W ., Fischer, G. W ., & Salovey, P. (1989). Perceived distributions of the characteristics of in-group and out-group members: Empirical evidence and a computer simulation. Journal o f Personality and Social Psychology, 57, 165-188. Linville, P. W ., Salovey, P., & Fischer, G. W. (1986). Stereotyping and perceived distributions of social characteristics: An application to in-group-out-group perception. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 165-208). Orlando, FL: Aca­ demic Press. Lippmann, W. (1922). Public opinion. New York: Harcourt, Brace, Jovanovitch. Locksley, A., Borgida, E ., Brekke, N ., & Hepburn, C. (1980). Sex stereotypes and social judg­ ment. Journal o f Personality and Social Psychology, 39, 821-831. Lord, C. G., & Saenz, D. S. (1985). Memory deficits and memory surfeits: Differential cognitive consequences of tokenism for tokens and observers. Journal o f Personality and Social Psycholo­ gy, 49, 918-926. Maass, A., & Schaller, M. (1991). Intergroup biases and the cognitive dynamics of stereotype for­ mation. In W. Stroebe & M. Hewstone (Eds.), European review o f social psychology (Vol. 2, pp. 189-209). Chichester, England: Wiley. Mackie, D. M ., Allison, S. T ., Worth, L. T., & Asuncion, A. G. (1992a). The generalization of outcome-biased counter-stereotypic inferences. Journal o f Experimental Social Psychology, 28, 43-64. Mackie, D. M ., Allison, S. T ., Worth, L. T ., & Asuncion, A. G. (1992b). The impact of outcome biases on counterstereotypic inferences about groups. Personality and Social Psychology Bulletin, 18, 44-51. Mackie, D. M ., Asuncion, A. G., & Rosselli, F. (1992). The impact of positive affect on persuasion processes. In M. S. Clark (Ed.), Review o f personality and social psychology (Vol. 14, pp. 247-270). Newbury Park, CA: Sage. Mackie, D. M ., & Hamilton, D. L. (1993). Affect, cognition, and stereotyping: Interactive process­ es in group perception. San Diego, CA: Academic Press. Mackie, D. M ., Hamilton, D. L., Schroth, H. A., Carlisle, C. J., Gersho, B. F., Meneses, L. M., Nedler, B. F., & Reichel, L. D. (1989). The effects of induced mood on illusory correlations. Journal o f Experimental Social Psychology, 25, 524-544. Mackie, D. M ., Sherman, J. W ., & Worth, L. T. (1993). On-line and memory-based processes in group variability judgments. Social Cognition, 11, 44-69. Mackie, D. M ., & Worth, L. T. (1991). “Feeling good but not thinking straight”: Positive mood and persuasion. In J. P. Forgas (Ed.), Emotion and social judgm ents (pp. 201-220). Oxford: Pergamon. Mackie, D. M ., Worth, L. T ., & Allison, S. T. (1990). Outcome biased inferences and the percep­ tion of change in groups. Social Cognition, 8, 325-342. Macrae, C. N., Hewstone, M ., & Griffiths, R. J. (1993). Processing load and memory for stereotypebased information. European Journal o f Social Psychology, 23, 77-87. Macrae, C. N., Milne, A. B., & Bodenhausen, G. V. (in press). Stereotypes as energy-saving devices: A peek inside the cognitive toolbox. Journal o f Personality and Social Psychology. Macrae, C. N., & Shepherd, J. W. (1989). Stereotypes and social judgments. British Journal o f So­ cial Psychology, 28, 319-325. Macrae, C. N., Shepherd, J. W ., & Milne, A. B. (1992). The effects of source credibility on the dilution of stereotype-based judgments. Personality and Social Psychology Bulletin, 18, 765-775.

64

HAMILTON AND SHERMAN

Major, B., & Crocker, J. (1993). Social stigma: The consequences of attributional ambiguity. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 345-370). San Diego, CA: Academic Press. Malt, B. C. (1989). An on-line investigation of prototype and exemplar strategies in classification. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 15, 539-555. Medin, D. L., Altom, M. W ., & Murphy, T. D. (1984). Given versus induced category representa­ tions: Use of prototype and exemplar information in classification. Journal o f Experimental P sy­ chology: Learning, Memory, and Cognition, 10, 333-352. Medin, D. L., & Schaffer, M. M. (1978). Context theory of classification learning. Psychological Review, 85, 207-238. Medin, D. L., & Wattenmaker, W. D. (1987). Category cohesiveness, theories, and cognitive ar­ chaeology. In U. Neisser (Ed.), Concepts and conceptual development (pp. 25-62). Cambridge, England: Cambridge University Press. Messick, D. M ., & Mackie, D. M. (1989). Intergroup relations. In M. R. Rosenzweig & L. W. Porter (Eds.), Annual review o f psychology (Vol. 40, pp. 45-81). Palo Alto, CA: Annual Reviews. Miller, N., & Brewer, M. B. (1986). Categorization effects on in-group and out-group perception. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 209-230). Orlando, FL: Academic Press. Mullen, B., & Hu, L. (1989). Perceptions of in-group and out-group variability: A meta-analytic integration. Basic and A pplied Social Psychology, 10, 233-252. Mullen, B., & Johnson, C. (1990). Distinctiveness-based illusory correlations and stereotyping: A meta-analytic integration. British Journal o f Social Psychology, 29, 11-28. Murphy, G. L., & Medin, D. L. (1985). The role of theories in conceptual coherence. Psychological Review, 92, 289-316. Neuberg, S. L., & Fiske, S. T. (1987). Motivational influences on impression formation: Outcome dependency, accuracy-driven attention, and individuating processes. Journal o f Personality and Social Psychology, 53, 431-444. Nosofsky, R. M. (1987). Attention and learning processes in the identification and categorization of integral stimuli. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 13, 87-108. Oakes, P. J., & Turner, J. C. (1990). Is limited information processing capacity the cause of social stereotyping? In W. Stroebe & M. Hewstone (Eds.), European review o f social psychology (Vol. 1, pp. 111-135). Chichester, England: Wiley. Ostrom, T. M., & Sedikides, C. (1992). Out-group homogeneity effects in natural and minimal groups. Psychological Bulletin, 112, 536-552. Park, B., & Hastie, R. (1987). Perception of variability in category development: Instance- versus abstraction-based stereotypes. Journal o f Personality and Social Psychology, 53, 621-635. Park, B., & Judd, C. M. (1990). Measures and models of perceived group variability. Journal o f Personality and Social Psychology, 59, 173-191. Park, B., Judd, C. M ., & Ryan, C. S. (1991). Social categorization and the representation of varia­ bility information. In W. Stroebe & M. Hewstone (Eds.), European review o f social psychology (Vol. 2, 211-245). Chichester, England: Wiley. Park, B., & Rothbart, M. (1982). Perception of out-group homogeneity and levels of social categori­ zation: Memory for the subordinate attributes of in-group and out-group members. Journal o f P er­ sonality and Social Psychology, 42, 1051-1068. Park, B., Ryan, C. S., & Judd, C. M. (1992). The role of meaningful subgroups in explaining differ­ ences in perceived variability for in-groups and out-groups. Journal o f Personality and Social Psy­ chology, 63, 553-567. Pettigrew, T. F. (1981). Extending the stereotype concept. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 303-332). Hillsdale, NJ: Lawrence Erlbaum Associates. Posner, M. I., & Keele, S. W. (1968). On the genesis of abstract ideas. Journal o f Experimental Psychology, 77, 353-363.

1.

STEREOTYPES

65

Pryor, J. B. (1986). The influence of different encoding sets upon the formation of illusory correla­ tions and group impressions. Personality and Social Psychology Bulletin, 72, 216-226. Quattrone, G. A. (1986). On the perception of a group’s variability. In S. Worchel & W. Austin (Eds.), The psychology o f intergroup relations (2nd ed., pp. 25-48). Chicago: Nelson-Hall. Quattrone, G. A., & Jones, E. E. (1980). The perception of variability within in-groups and out­ groups: Implications for the Law of Small Numbers. Journal o f Personality and Social Psycholo­ gy, 38, 141-152. Reeder, G. D., & Brewer, M. B. (1979). A schematic model of dispositional attribution in interper­ sonal perception. Psychological Review, 86, 61-79. Rojahn, K., & Pettigrew, T. F. (1992). Memory for schema-relevant information: A meta-analytic resolution. British Journal o f Social Psychology, 31, 81-110. Rosch, E. H. (1978). Principles of categorization. In E. Rosch & B. B. Lloyd (Eds.), Cognition and categorization (pp. 27-48). Hillsdale, NJ: Lawrence Erlbaum Associates. Rothbart, M. (1981). Memory processes and social beliefs. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 145-182). Hillsdale, NJ: Lawrence Erlbaum Associates. Rothbart, M ., Evans, M ., & Fulero, S. (1979). Recall for confirming events: Memory processes and the maintenance of social stereotyping. Journal o f Experimental Social Psychology, 15, 343-355. Rothbart, M ., Fulero, S., Jensen, C., Howard, J., & Birrell, B. (1978). From individual to group impressions: Availability heuristics in stereotype formation. Journal o f Experimental Social P sy­ chology, 14, 237-255. Rothbart, M., & John, O. P. (1985). Social categorization and behavioral episodes: A cognitive analysis of the effects of intergroup contact. Journal o f Social Issues, 47(3), 81-104. Rothbart, M ., & Lewis, S. (1988). Inferring category attributes from exemplar attributes: Geometric shapes and social categories. Journal o f Personality and Social Psychology, 55, 861-872. Rothbart, M., & Park, B. (1986). On the confirmability and disconfirmability of trait concepts. Journal o f Personality and Social Psychology, 50, 131-142. Sagar, H. A., & Schofield, J. W. (1980). Racial and behavioral cues in black and white children’s perceptions of ambiguously aggressive acts. Journal o f Personality and Social Psychology, 39, 590-598. Sanbonmatsu, D. M ., Shavitt, S., Sherman, S. J., & Roskos-Ewoldsen, D. R. (1987). Illusory corre­ lation in the perception of performance by self or a salient other. Journal o f Experimental Social Psychology, 23, 518-543. Sanbonmatsu, D. M ., Sherman, S. J., & Hamilton, D. L. (1987). Illusory correlation in the percep­ tion of individuals and groups. Social Cognition, 5, 1-25. Schaller, M. (1991). Social categorization and the formation of group stereotypes: Further evidence for biased information processing in the perception of group-behavior correlations. European Journal o f Social Psychology, 21, 25-35. Schaller, M ., & Maass, A. (1989). Illusory correlation and social categorization: Toward an integra­ tion of motivational and cognitive factors in stereotype formation. Journal o f Personality and So­ cial Psychology, 56, 709-721. Schwarz, N. (1990). Feelings as information: Informational and motivational functions of affective states. In R. Sorrentino & E. T. Higgins (Eds.), Handbook o f motivation and cognition: Founda­ tions o f social behavior (Vol. 2, pp. 527-561). New York: Guilford. Sherif, M. (1967). Group conflict and cooperation. London: Routledge & Kegan Paul. Sherman, J. W ., & Hamilton, D. L. (in press). On the formation of interitem links in person memory. Journal o f Experimental Social Psychology.

Simon, B., & Brown, R. (1987). Perceived intragroup homogeneity in minority-majority contexts. Journal o f Personality and Social Psychology, 53, 703-711. Simon, B., & Hamilton, D. L. (in press). Self-stereotyping and social context: The effects of relative in-group size and in-group status. Journal o f Personality and Social Psychology. Skov, R. B., & Sherman, S. J. (1986). Information-gathering processes: Diagnosticity, hypothesisconfirmatory strategies, and perceived hypothesis confirmation. Journal o f Experimental Social Psychology, 22, 93-121.

66

HAMILTON AND SHERMAN

Slusher, M. P., & Anderson, C. A. (1987). When reality monitoring fails: The role of imagination in stereotype maintenance. Journal o f Personality and Social Psychology, 52, 653-662. Smith, E. R. (1990). Content and process specificity in the effects of prior experiences. In T. K. Srull & R. S. Wyer, Jr. (Eds.), Advances in social cognition (Vol. 3, pp. 1-59). Hillsdale, NJ: Lawrence Erlbaum Associates. Smith, E. R. (1991). Illusory correlation in a simulated exemplar-based memory. Journal o f Experimen­ tal Social Psychology, 27, 107-123. Smith, E. R., & Zarate, M. A. (1990). Exemplar and prototype use in social categorization. Social Cognition, 8, 243-262. Smith, E. R., & Zarate, M. A. (1992). Exemplar-based model of social judgment. Psychological Review, 99, 3-21. Snyder, M ., & Swann, W. B. (1978). Hypothesis-testing processes in social interaction. Journal o f Personality and Social Psychology, 36, 1202-1212. Snyder, M ., Tanke, E. D., & Berscheid, E. (1977). Social perception and interpersonal behavior: On the self-fulfilling nature of social stereotypes. Journal o f Personality and Social Psychology, 35, 656-666. Srull, T. K. (1981). Person memory: Some tests of associative storage and retrieval models. Journal o f Experimental Psychology: Human Learning and Memory, 7, 440-462. Srull, T. K., Lichtenstein, M ., & Rothbart, M. (1985). Associative storage and retrieval processes in person memory. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 11, 316-345. Srull, T. K., & Wyer, R. S., Jr. (1979). The role of category accessibility in the interpretation of information about persons: Some determinants and implications. Journal o f Personality and So­ cial Psychology, 37, 1660-1672. Srull, T. K., & Wyer, R. S., Jr. (1980). Category accessibility and social perception: Some implica­ tions for the study of person memory and interpersonal judgments. Journal o f Personality and So­ cial Psychology, 38, 841-856. Srull, T. K., & Wyer, R. S., Jr. (1989). Person memory and judgment. Psychological Review, 96, 58-83. Stangor, C., & Duan, C. (1991). Effects of multiple task demands upon memory for information about social groups. Journal o f Experimental Social Psychology, 27, 357-378. Stangor, C., & Ford, T. E. (1992). Accuracy and expectancy-confirming processing orientations and the development of stereotypes and prejudice. In W. Stroebe & M. Hewstone (Eds.), Europe­ an review o f social psychology (Vol. 3, pp. 57-89). Chichester, England: Wiley. Stangor, C., & McMillan, D. (1992). Memory for expectancy-congruent and expectancy-incongruent information: A review of the social and social developmental literatures. Psychological Bulletin, 111, 42-61. Stangor, C., & Ruble, D. N. (1989). Stereotype development and memory: What we remember de­ pends on how much we know. Journal o f Experimental Social Psychology, 25, 18-35. Stangor, C., Sullivan, L. A., & Ford, T. E. (1991). Affective and cognitive determinants of prejudice. Social Cognition, 9, 359-380. Stephan, W. G. (1985). Intergroup relations. In G. Lindzey & E. Aronson (Eds.), The handbook o f social psychology (3rd ed., Vol. 2, pp. 599-658). New York: Random House. Stephan, W. G., & Stephan, C. W. (1985). Intergroup anxiety. Journal o f Social Issues, 47(3), 157-175. Stephan, W. G., & Stephan, C. W. (1993). Cognition and affect in stereotyping: Parallel interactive networks. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Inter­ active processes in group perception (pp. 111-136). San Diego, CA: Academic Press. Stern, L. D., Marrs, S., Millar, M. G., & Cole, E. (1984). Processing time and the recall of incon­ sistent and consistent behaviors of individuals and groups. Journal o f Personality and Social P sy­ chology, 47, 253-262. Stroebe, W ., & Insko, C. A. (1989). Stereotyping, prejudice, and discrimination: Changing concep­ tions in theory and research. In D. Bar-Tal, C. F. Graumann, A. W. Kruglanski, & W. Stroebe (Eds.), Stereotypes and prejudice: Changing conceptions (pp. 3-36). New York: Springer-Verlag.

1.

STEREOTYPES

67

Stroessner, S. J. (1992). Target-based determinants o f social categorization: Extensions o f the "white male norm”hypothesis. Unpublished doctoral dissertation, University of California, Santa Barbara. Stroessner, S. J., Hamilton, D. L., & Mackie, D. M. (1992). Affect and stereotyping: The effect of induced mood on distinctiveness-based illusory correlations. Journal o f Personality and Social Psychology, 62, 564-576. Stroessner, S. J., & Mackie, D. M. (1992). The impact of induced affect on the perception of varia­ bility in social groups. Personality and Social Psychology Bulletin, 18, 546-554. Stroessner, S. J., & Mackie, D. M. (1993). Affect and perceived group variability: Implications for stereotyping and prejudice. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 63-86). San Diego, CA: Academic Press. Swann, W. B., Jr. (1984). Quest for accuracy in person perception: A matter of pragmatics. Psycho­ logical Review, 91, 457-477. Tajfel, H. (1969). Cognitive aspects of prejudice. Journal o f Social Issues, 25(4), 79-97. Tajfel, H. (1970). Experiments in intergroup discrimination. Scientific American, 223(5), 96-102. Tajfel, H., Billig, M ., Bundy, R. P., & Flament, C. (1971). Social categorization and intergroup behavior. European Journal o f Social Psychology, 1 , 149-177. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology o f intergroup relations (pp. 33-47). Monterey, CA: Brooks/Cole. Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel & W. Austin (Eds.), The psychology o f intergroup relations (2nd ed., pp. 7-24). Chicago: Nelson-Hall. Tajfel, H., & Wilkes, A. L. (1963). Classification and qualitative judgment. British Journal o f Psy­ chology, 54, 101-114. Taylor, S. E. (1981). A categorization approach to stereotyping. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 88-114). Hillsdale, NJ: Lawrence Erlbaum Associates. Taylor, S. E., & Crocker, J. (1981). Schematic bases of social information processing. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario Symposium (Vol. 1, pp. 89-134). Hillsdale, NJ: Lawrence Erlbaum Associates. Taylor, S. E., Fiske, S. T., Etcoff, N. L., & Ruderman, A. (1978). Categorical bases of person memory and stereotyping. Journal o f Personality and Social Psychology, 36, 778-793. Turner, J. C. (1987). Rediscovering the social group: A self-categorization theory. New York: Basil Blackwell. Tversky, A. (1977). Features of similarity. Psychological Review, 84, 327-352. Tversky, A., & Kahneman, D. (1973). Availability: A heuristic forjudging frequency and probabili­ ty. Cognitive Psychology, 5, 207-232. Vanbeselaere, N. (1991). The different effects of simple and crossed categorizations: A result of the category differentiation process or of differential category salience? In W. Stroebe & M. Hewstone (Eds.), European review o f social psychology (Vol. 2, pp. 247-278). Chichester, England: Wiley. Vanman, E. J., & Miller, N. (1993). Applications of emotion theory and research to stereotyp­ ing and intergroup relations. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 213-238). San Diego, CA: Academic Press. Weber, R., & Crocker, J. (1983). Cognitive processes in the revision of stereotypic beliefs. Journal o f Personality and Social Psychology, 45, 961-977. Wilder, D. A. (1978). Homogeneity of jurors: The majority’s influence depends upon their perceived independence. Law and Human Behavior, 2, 363-376. Wilder, D. A. (1981). Perceiving persons as a group: Categorization and intergroup relations. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 213-258). Hillsdale, NJ: Lawrence Erlbaum Associates.

68

HAMILTON AND SHERMAN

Wilder, D. A. (1993). The role of anxiety in facilitating stereotypic judgment of out-group behavior. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 87-109). San Diego, CA: Academic Press. Wilder, D. A., & Shapiro, P. (1989). Role of competition-induced anxiety in limiting the beneficial impact of positive behavior by an out-group member. Journal o f Personality and Social Psycholo­ g y , 56, 60-69. Word, C. O.^ Zanna, M. P., & Cooper, J. (1974). The nonverbal mediation of self-fulfilling prophe­ cies in interracial interaction. Journal o f Experimental Social Psychology, 10, 109-120. Zadny, J., & Gerard, H. B. (1974). Attributed intentions and informational selectivity. Journal o f Experimental Social Psychology, 10, 34-52. Zarate, M. A., & Smith, E. R. (1990). Person categorization and stereotyping. Social Cognition, 8 , 161-185.

2

Cognitive Processes in Attitude Change

Richard E. Petty Joseph R. Priester Duane T. W egener The Ohio State University

Contents The Nature and Properties of Attitudes and Attitude Systems Underlying Bases of Attitudes 71 Attitude and Belief Structure 75 Attitude Strength 80 Information Processing Stages in Attitude Change 81 Reception/Learning Processes 82 Evaluation/Yielding Processes 85 Cognitive Response Approach 85 The Elaboration Likelihood Model o f Persuasion 88 Peripheral Mechanisms in Attitude Change 104 Information Integration Stage 110 Consequences of Different Attitude Change Processes 112 Persistence of Attitude Change 113 Resistance to Counterpersuasion 114 Attitude-Behavior Consistency 116 Summary and Conclusions 118 Acknowledgments 119 References 120

69

THE NATURE AND PROPERTIES OF ATTITUDES AND ATTITUDE SYSTEMS In the 1992 presidential election, public opinion polls showed large swings in people’s attitudes toward the candidates over relatively short time periods. About a year before the actual election, the incumbent President George Bush was held in high esteem by about 80% o f the electorate. By the time o f the Democrat­ ic National Convention, his approval rating plunged to around 30%, and his

69

70

PETTY, PRIESTER, WEGENER

opponent, Arkansas Governor Bill Clinton, was ahead in the polls by nearly 30 points. By the end of the Republican National Convention, Clinton’s lead over Bush had slipped to around 10 points. At this point, the polls finally became some­ what more stable. What psychological processes account for the various dramat­ ic changes in attitudes that were observed prior to the election? The goal of this chapter is to explicate the various psychological processes that are responsible for attitude change. A common definition of attitude is a general and relatively enduring evalua­ tion of some person (including oneself), group, object, or issue (Petty & Cacioppo, 1986a). By relatively enduring, we simply mean that the evaluation associated with the attitude object is stored in long-term memory, and therefore is potential­ ly retrievable (e.g., Fazio, 1986). By general, we mean that an attitude refers to the person’s most global or overall assessment of the attitude object. Specific experiences and attributes associated with the attitude object also are stored in memory and can vary in their individual evaluations. Although psychologists typi­ cally measure attitudes on a bipolar evaluative continuum (e.g., Osgood, Suci, & Tannenbaum, 1957), this does not mean that attitudes necessarily are represented in a scalar fashion in memory (e.g., Devine & Ostrom, 1988), or that additional information about attitudes cannot be gleaned by assessing the positive and nega­ tive features of attitudes independently (e.g., Kaplan, 1972; Thompson, Zanna, & Griffin, in press). Attitudes typically are assumed to be learned over time, although it is now clear that there are at least indirect genetic contributions to attitude development as well (e.g., Eaves, Eysenck, & Martin, 1989; Tesser, 1993). In contrast to the traditional view of attitudes as stored in long-term memory, some attitude theorists have taken the position that people’s evaluations are not particularly stable entities and that attitudes might be constructed each time they are needed on the basis of whatever information is salient (see Tesser, 1978; Wil­ son & Hodges, 1992; Wyer & Srull, 1989). Zaller (1992) argued that many po­ litical attitudes were described best in this manner. Although some longitudinal studies found considerable stability in attitudes over many years (e.g., Marwell, Aiken, & Demerath, 1987), others found that people’s attitudinal reports can be influenced by such transient factors as their current mood states (see Schwarz, 1990), their most recent behaviors (see Bern, 1972; Fazio, Herr, & Olney, 1984), and seemingly minor variations in the manner by which they are asked an atti­ tude question (see Tourangeau & Rasinski, 1988). Our view is that evaluations are not invariably or even typically constructed anew each time an attitude object is confronted (see also Lingle & Ostrom, 1981). When people are asked about their attitudes, what happens? First, the mere men­ tion of the attitude object or issue can make an attitude spontaneously accessible (Fazio, Sanbonmatsu, Powell, & Kardes, 1986). In fact, recent research sug­ gests that attitudes toward a wide variety of objects are retrieved automatically upon the mere presentation of the attitude object (Bargh, Chaiken, Govender,

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

71

& Pratto, 1992). However, if no attitude is immediately accessible, the person might search memory for a stored evaluation if motivated and able to do so. If no evaluation is found, or if the person is unwilling to report the attitude retrieved (e.g., due to insufficient confidence in it), the person might construct an attitude based on accessible information from memory and in the current context. Before turning to the various specific theories of attitude change, it is useful to review some work on the structure and function of attitudes (see also Pratkanis, Breckler, & Greenwald, 1989). To the extent that we understand what the underlying bases of attitudes are and their structure, we presumably can under­ stand what can be targeted to change attitudes. Following this, we discuss the various processes and theories of attitude change that guide contemporary research. We conclude by outlining the consequences of these processes. Underlying Bases of Attitudes

Beliefs, Emotions, and Behaviors as Bases o f Attitudes, Although an attitude refers to a general evaluation of some object, this evaluation can be based on various beliefs (i.e., specific attributes associated with the attitude object such as “Candidate A is tall” or “favors capital punishment”), emotions (affective states and experiences such as “Candidate A makes me happy”), and/or behaviors (con­ crete actions such as “I put a sign for Candidate A in my yard”). Also, an attitude can influence these three categories of responses (Petty & Cacioppo, 1986b; Zanna & Rempel, 1988). That is, a person might come to favor a presidential candidate only after being induced to contribute money to the cause, after finding the planks in the party platform to be compelling, or after feeling a warm pride in response to patriotic images featured in a television commercial for the candidate. Simi­ larly, a person who already evaluates a candidate favorably might engage in the behavior of endorsing the candidate to friends, might process the presidential de­ bates in a biased fashion, and might feel happiness when the candidate wins the election. Research suggests that beliefs, emotions, and behaviors all can contribute separately to people’s attitudes (e.g., Abelson, Kinder, Peters, & Fiske, 1982; Breckler, in press; Stangor, Sullivan, & Ford, 1991), although in some em­ pirical studies it is difficult to separate the variance contributed by the individual components from the procedures employed to measure them (Crites, Fabrigar, & Petty, 1993). Although an assumption of the “tripartite” model of attitudes was that attitudes tended to be composed of all three categories of responses (e.g., Rosenberg & Hoviand, 1960), more recent research has emphasized the notion that attitudes can be based on just one or two of the components (e.g., Millar & Tesser, 1986b). Similarly, although earlier research tended to emphasize a strain toward consistency among the components of attitudes (especially the emo­ tional and belief components; e.g., Rosenberg, 1960), recent research has begun to emphasize the implications of inconsistency and ambivalence among the various

72

PETTY, PRIESTER, WEGENER

bases (e.g., Chaiken, Pomerantz, & Giner-Sorolla, in press; Millar & Tesser, 1992; Thompson et al., in press; Wilson, Dunn, Kraft, & Lisle, 1989). Current models of attitude structure recognize that each of the so-called components can be consistent or inconsistent with the attitude (i.e., one can examine attitudebehavior consistency, attitude-belief consistency, and attitude-emotion consisten­ cy), and that the attitudinal bases therefore can be consistent or inconsistent with each other.1Resolution of inconsistency likely requires some cognitive effort so that attitudes that are thought about more frequently are likely to exhibit greater component consistency (Breckler & Wiggins, 1989a). Some theorists have proposed that the affective qualities of an attitude are more fundamental or take precedence over the cognitive bases of the evaluation. For example, Zajonc (1980) and Zajonc and Markus (1984) argued that affect typi­ cally precedes cognition in the formation of attitudes. In a similar vein, Katz and Stotland (1959) argued that an inconsistency between affect and cognition is resolved more readily by changing beliefs, rather than the emotions associated with the object. Breckler and Wiggins (1991) found that the initial affect associated with an issue was a better predictor of the thoughts generated to a message on this topic than were the initial beliefs associated with the topic. However, be­ cause so little research has addressed these issues explicitly, confident general conclusions are premature. Functional Bases o f Attitudes. Although the tripartite model of attitudes iden­ tifies three categories of attributes that can form the basis of an attitude, it is not clear what kinds of beliefs, emotions, or behaviors will be the most important. Katz (1960) and Smith, Bruner, and White (1956) addressed this issue in terms of the psychological needs that an attitude can serve for a person. Perhaps the most fundamental function of attitudes is the knowledge or object appraisal func­ tion. That is, virtually all attitudes assist the person in understanding and making sense of the world. Quickly retrieving evaluations that tell individuals whether objects encountered are good (safe) or bad (threatening) makes everyday life easier by minimizing the need to assess and construct these evaluations effortfully each time the attitude object is encountered (Cacioppo, Petty, & Bemtson, 1991; Smith et al., 1956). Because attitudes provide a ready evaluation, they also reduce the stress of decisionmaking (Blascovich et al., 1993; Fazio, Blascovich, & Dris­ coll, 1992). In addition to this basic appraisal function, some attitudes are thought to serve more specific motives. For example, some attitudes are postulated as protecting people from threatening truths about themselves or serving to enhance self-image (i.e., an ego-defensive function). Still other attitudes are thought to be based on Although Rosenberg (1956) developed the notion of affective-cognitive consistency, in the cur­ rent terminology he was examining attitude-belief consistency (or what Chaiken et al., in press, call evaluative-cognitive consistency).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

73

the extent to which they lead to explicit rewards and/or punishments (utilitarian function) or give expression to important values (value-expressive function). After a period of neglect, persuasion researchers are beginning to show renewed interest in functional theories. Researchers lost interest in functional approaches, because it did not appear to be possible to assess what specific functions attitudes held for people (Kiesler, Collins, & Miller, 1969). At present, several potential solutions to this problem have been suggested. One solution relies on individual differences and suggests that attitudes serve different functions for different kinds of people. For example, Snyder and DeBono (1985) hypothesized that the atti­ tudes of low “self-monitors” (see Snyder, 1979) would serve primarily a valueexpressive function, whereas the attitudes of high self-monitors would serve primarily a social adjustive function (e.g., adopting attitudes that provide rewards from valued peers). An alternative to the personality approach proposes that many issues and objects serve a common function for a wide variety of people (e.g., Abelson & Prentice, 1989; Prentice, 1987; Shavitt, 1989). For example, atti­ tudes toward air conditioners probably serve a utilitarian function for most peo­ ple. In a third approach, Herek (1987) employed the thought-listing procedure (see Cacioppo & Petty, 1981b) to analyze the functional bases of attitudes on an individual level. Despite the promising new directions in functional theory, some pitfalls re­ main. For instance, the Herek (1987) procedure allows functions to be assessed post hoc, which is useful for understanding the basis of individuals’ attitudes and designing individual persuasion treatments, but is amenable to a mass communi­ cation context only if the functional basis of an attitude is widespread. However, it seems unlikely that most attitude objects would serve the same function for most people (cf., Shavitt, 1989). The general personality strategy of Snyder and DeBono (1989) could be useful, but only if research evidence revealed that other personality variables are linked consistently to particular functions (e.g., for high authoritarians, attitudes serve an ego-defensive function; Snyder & DeBono, 1989). Changing Attitudes with Different Bases. Research indicates that some peo­ ple find certain types of arguments and experiences more persuasive than others. For example, in one study, students at a university with a religious affiliation were divided into those who had a “religious” or a “legalistic” self-schema (Caciop­ po, Petty, & Sidera, 1982). People with a religious self-schema were those who endorsed religious self-descriptions more quickly than legalistic ones (and vice versa for the legalistic subjects; see Markus, 1977). When these individuals were presented with persuasive arguments on the topic of capital punishment that made either a religious or a legalistic appeal, they found the arguments that matched their self-schemas to be more persuasive. In another study, rather than relying on chronically important dimensions of judgment, particular dimensions were primed (Sherman, Mackie, & Driscoll, 1990). For example, subjects were asked to evaluate political candidates who had positive attributes on one dimension (e.g.,

74

PETTY, PRIESTER, WEGENER

foreign policy) and negative attributes on another (e.g., the economy). Prior to the judgment, subjects were required to memorize a list of words that either primed the foreign policy dimension (e.g., overseas) or the economic dimension (e.g., treasury). Individuals for whom a particular dimension had been primed were more likely to base their evaluations of the candidates on information relevant to the primed dimension than on the other equally relevant, but less accessible, dimension (see also Iyengar, Kinder, Peters, & Krosnick, 1984). Thus, one can increase the effectiveness of a persuasive appeal by presenting favorable infor­ mation relevant to dimensions of judgment that are either chronically or tem­ porarily salient for an individual. Would a similar matching strategy be effective when considering the underly­ ing bases of attitudes? Presumably, the bases of attitudes reflect the dimensions that are important and salient for individuals. For example, for some people (or objects), the central merits of the object might be mostly affective in nature (i.e., how the object makes them feel), and these feelings might stem from concerns about social relations (i.e., a social adjustment function). To change such an atti­ tude, should the person be provided with new arguments and experiences that match the basis of the attitude (i.e., challenging the affectively based attitude with affect; providing arguments relevant to social adjustment concerns), or would a mismatching strategy be more successful? Interestingly, in studies examining the affective (emotional) versus cognitive (belief) bases of attitudes, some inves­ tigators have found evidence favoring a matching strategy—especially for affec­ tively based attitudes (Edwards, 1990), whereas others have obtained evidence favoring mismatching (Millar & Millar, 1990). However, with respect to the func­ tional foundations of attitudes, research consistently has supported the matching principal. For example, in several studies, high and low self-monitors were found to be more susceptible to arguments that matched the presumed functional basis of their attitudes. That is, high self-monitors were more susceptible to arguments based on the image of a consumer product than were low self-monitors (Snyder & DeBono, 1987, 1989; see also DeBono, 1987).2 Matching arguments to the functions inherent in attitude objects also has been successful (Shavitt, 1990). In general, whether matching or mismatching strategies are more effective should depend on a number of factors. One consideration involves the strength and relevance of the challenging stimulus. That is, does the new information ef­ fectively undermine the evidential or functional basis of the initial beliefs or does it simply present additional inconsistent information (see Petty, Gleicher, & Baker, 1991)? For example, if it is implausible to attack the utilitarian basis of liking for an object (i.e., no strong utilitarian arguments are available), one could indi­ 2DeBono and Hamish (1988) also employed a functional analysis to predict which sources would affect information processing for high and low self-monitors. Specifically, they proposed and found evidence for the view that an attractive source would engender more interest and enhance message processing for the image-oriented high self-monitors, but that an expert source would enhance processing for the more value-expressive low self-monitors.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

75

cate that the object is linked negatively to social adjustment concerns. In addition to the strength of the arguments available, the strength of the attitude under at­ tack also could be an important consideration (e.g., is the attitude well estab­ lished or newly formed? See Olson & Zanna, 1993). For example, a person’s initial attitude might be based on whatever information is available at the time a judgment is required if information central to the merits of the object is un­ available. Such incidental attitudes might be changed easily by subsequently providing information on more fundamental dimensions, mismatching the origi­ nal basis of the attitude. On the other hand, if the original attitude is based on more central information, a matching strategy might be required. Attitude and Belief Structure

Current models of attitudes suggest that an attitude is linked in memory to vari­ ous beliefs, emotions, and/or behaviors, and that the attitude and the information supporting it serve certain functions (e.g., allowing the person to understand the world, feel good about oneself, etc.). In addition to the very general structural idea that individual attitudes are linked to beliefs, emotions, and behaviors, some theorists have addressed more particular structural notions, especially as they re­ late to the structure of the beliefs that comprise individual attitudes. We describe three such models next. Probabilogical Model. Some theorists have conceptualized attitudes as the end result of a syllogistic network of beliefs (Bern, 1970; McGuire, 1960a; Wyer, 1970, 1974). For example, at a minimum, an attitude can be based on two premises that lead to a conclusion: Premise 1 = Candidate A favors gun control; Premise 2 = Gun control is good for the country; Conclusion = Candidate A is good for the country. The conclusions of attitudinal syllogisms such as this can vary in the extent to which they have an extensive horizontal and vertical structure. This syllogism would have an extensive vertical structure to the extent that its premises were derived from many other logical syllogisms. The syllogism would have an extensive horizontal structure to the extent that its conclusion could be derived from many other logical syllogisms (McGuire, 1981). McGuire (1960a, 1960b) and Wyer (1970, 1974) developed mathematical models of belief syllogisms. For example, assume that you think that it is highly likely (p = .9) that any candidate who favors gun control would be good for the country, but you are unsure of whether Candidate A is in favor of gun con­ trol (p = .5). How confident would you be that Candidate A is good for the coun­ try? Consider the following reworking of the earlier syllogism: Premise 1:

If Candidate A favors gun control, then Candidate A is good for the country [p(B/A)] = .9 Premise 2: Candidate A favors gun control [p(A)] = .5 Conclusion: Candidate A is good for the country [p(B)] = ?

76

PETTY, PRIESTER, WEGENER

Using probability theory, Wyer (1970) noted that: p(B) = p(B/A)*p(A) + p{BI~A)*p{~A)

(1)

To calculate p(B), we need to know the probability that Candidate A does not favor gun control, or p(~A). Because a candidate only can favor or disfavor gun control, the probability of (~A) = 1 - p(A) = .5. We also need to know the probability that you would support Candidate A even if he did not favor gun con­ trol, or p(BI~A). Let us assume that the latter is only .3. Using Equation 1, it is predicted that: p{B) = (.9)*(.5) + (.3)*(.5) = .60.

(2)

If people’s beliefs are organized in these logical structures and people reason logically, we can use the formula to predict how new information can change a person’s beliefs. For example, if Candidate A gave a speech that led you to believe that there was a .8 chance that he favored gun control (and thus .2 that he did not), substituting these values into Equation 1 produces the prediction that you should believe more strongly that Candidate A is good for the country. That is: p(B) = (.9)*(.8) + (.3)*(.2) = .78

(3)

Your belief that Candidate A is good for the country has increased from .6 to .78.3 A number of studies testing the viability of the notion that new information changes people’s beliefs in the manner predicted by the probabilogical model have supported its utility (Wyer, Carlston, & Hartwick, 1979; Wyer & Goldberg, 1970). Of course, the fact that Equation 1 provides accurate predictions does not mean that beliefs actually are organized into syllogistic structures or that people retrieve beliefs from these structures to evaluate conclusions. In fact, Wyer and Hart­ wick (1980) viewed Equation 1 as simply describing an inference process in which people compute their beliefs in the conclusion (B) by considering whatever previ­ ously acquired piece of relevant information comes to mind most easily (e.g., A). The piece of information that comes to mind could be a belief that is linked syllogistically to the conclusion in memory, or it could be a belief that is momen­ tarily salient due to situational factors. Wyer and Hartwick argued that, in this inference process, people first consider the implications of A for B. Then, they consider the likelihood of B if A were not true. Finally, if these two inferences are different, they average the two conditional probabilities, weighting them by their beliefs that A was and was not likely to occur, respectively. 3Note that the quantity [p{B/A) - p (B /-A )] is relevant to assessing the impact of a change in p{A) on p{B). That is, the larger the difference between p(B/A) and p(B /~A ), the greater the impact a change in belief A has on conclusion B (Wyer, 1970). Therefore, knowledge of these conditional probabilities would be useful in selecting specific beliefs to change (e.g., Candidate A favors gun control vs. Candidate A favors tax reductions) to produce maximal change in a desired conclusion (e.g., Candidate A is good for the country).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

77

Regardless of the specific interpretation of Equation 1, the probabilogical for­ mula demonstrates how a targeted change in one belief (e.g., Candidate A favors gun control) can have an impact on a completely unmentioned belief (e.g., Can­ didate A is good for the country). Consistent with this view, research has shown that a message targeted at changing abortion attitudes can produce changes in people’s attitudes toward the related, but unmentioned, issue of contraception (Mugny & Perez, 1991; see also, Dillehay, Insko, & Smith, 1966). In fact, merely responding to an attitude item on one issue can produce an attitude polarization effect on a related issue (Henninger & Wyer, 1976; Judd, Drake, Downing, & Krosnick, 1991). The more logical connections or associative links that exist among beliefs or that people perceive as they think about an issue, the greater the number of beliefs that could change as a result of a single belief change. Expectancy-Value Model. The probabilogical model outlined earlier considers only the probabilities that certain statements are true or likely (e.g., that Candi­ date A is linked to a trait such as gun control), or that the trait is linked to some other property (e.g., good for the country). It does not consider degrees of desir­ ability of the traits explicitly.4 Although the probabilogical model can be modi­ fied to incorporate variations in desirability (see Wyer, 1973), the most popular approach to considering degrees of likelihood and desirability comes from expectancy-value theories. In particular, expectancy-value theorists analyze atti­ tudes by focusing on the extent to which people expect the attitude object to be related to important values or produce positive versus negative consequences (e.g., Peak, 1955; Rosenberg, 1956; see reviews by Bagozzi, 1984, 1985). In one in­ fluential formulation, Fishbein and Ajzen (1975; Ajzen & Fishbein, 1980) hold that the attributes (or consequences) associated with an attitude object are evalu­ ated along two dimensions. That is, a person considers the likelihood (expectan­ cy) that an attribute or consequence is associated with the object along with the desirability (value) of that attribute or consequence. The desirability of each at­ tribute (e,) is weighted by the likelihood that the object possesses the attribute (b,)9 and the quantity is summed over all attributes. Stated mathematically, the attitude toward the object (Aa) is expressed in Equation 4: (A0) = I (V,)

(4)

Although questions occasionally arise about various aspects of this model (e.g., Cronen & Conville, 1975; Evans, 1991), a monumental body of research sup­ ports the idea that attitudes toward objects, issues, and people become more favor­ able as the number of likely desirable consequences (or attributes) and unlikely undesirable consequences associated with them increase (see Ajzen & Fishbein, 1980; Fishbein, 1980; Fishbein & Ajzen, 1975; and Pieters, 1988, for reviews). 4However, McGuire (1960a) noted that the perceived likelihood of a conclusion is influenced by the desirability of the premises and the conclusion. That is, beliefs are influenced by both “logical consistency” and “hedonic consistency.”

78

PETTY, PRIESTER, WEGENER

The major implication of expectancy-value theories for attitude change is that a persuasive message will be effective to the extent that it produces a change in either the likelihood or the desirability component of an attribute that is linked to the attitude object. For example, consider a person who, prior to message ex­ posure, believes that Candidate A has a .5 chance of favoring gun control—a consequence valued at 7 on a 10-point scale. The message will produce a more positive attitude toward Candidate A to the extent that it either increases the likeli­ hood that the candidate favors gun control beyond .5, or increases the desirabili­ ty of gun control beyond 7. Given the wide applicability of Fishbein and Ajzen’s approach, and the extensive number of studies documenting the link between at­ titudes and the likelihood and desirability components of beliefs, it is perhaps surprising that relatively little work on attitude change has been guided explicitly by this framework. Nevertheless, existing research supports the view that mes­ sages can influence attitudes by changing either the evaluation or the likelihood component of beliefs (e.g., Lutz, 1975; MacKenzie, 1986). Importantly, as might be expected given the thoughtful processing assumed by this approach, research indicates that the expectancy-value framework accounts for less variance in atti­ tudes when topic-relevant knowledge is low (e.g., Lutz, 1977). Thought Systems Approach. In a major elaboration of the probabilogical model, McGuire and McGuire (1991) addressed the distinctive origins of likeli­ hood and evaluation judgments. The McGuires noted that when thinking about “core events” such as “electing Candidate A,” people think about both the likeli­ hood and the desirability of the event. The likelihood of an event is postulated to be related to the number of antecedents of the event that a person can gener­ ate. That is, the event is seen as more likely the greater the number of conditions a person can think of that would bring the event about and less likely the greater the number of conditions that would prevent it. However, the desirability of an event is postulated to be determined by the consequences of the event. That is, the event is seen as more desirable the greater the number of good things that it would provoke and less desirable the greater the number of bad things that it would inspire. The desirability of the event also is postulated to be related to the perceived goodness of the antecedents of the event.5 The McGuires’ (1991) analysis of core events is applicable to understanding the strength of the arguments in a persuasive communication, especially when people think actively about the message (Petty & Wegener, 1991). For example, an argument in favor of electing Candidate A, such as “Candidate A will reduce taxes,” will be more persuasive to the extent that a person generates (or a mes­ sage explains) specific ways in which the candidate can bring about a reduction in taxes and provides few impediments to this action. Similarly, the argument will be more persuasive the greater the number of beneficial and the fewer the number of harmful consequences that the person generates about reducing taxes. 5The thought systems model also contains a number of other interesting postulates (see Wyer & Srull, 1991, for a complete review).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

79

Consistent with Fishbein and Ajzen’s (1975) expectancy-value model, Ajzen (1991) noted that, in addition to considering the number of antecedents and consequences, it would be useful to quantify the desirability and likelihood of the salient antece­ dents and consequences associated with an event. However, if people are not thinking actively about the message arguments, evaluation and likelihood judgments could be rendered by more simplistic analyses that are not considered by the thought systems approach. For example, Wyer (1991) suggested that the likelihood of an event can be judged simply on the basis of the perceived familiarity of the event where familiarity is determined by the number of event-related concepts that are accessible in memory. Consistent with the familiarity idea, Arkes, Boehm, and Xu (1991) showed that mere repetition of a statement led to an increase in its perceived truthfulness (see also, Hawkins & Hoch, 1992). In an interesting analysis of perceived validity, Gilbert (1991) argued that propo­ sitions are assumed to be true until or unless countervailing information is retrieved (i.e., until they are deemed untrue). Thus, disrupted processing during the presen­ tation of false statements should increase their perceived validity (Gilbert, Krull, & Malone, 1990; Gilbert, Tafarodi, & Malone, 1993). Gruenfeld and Wyer (1992) showed that denials of information an individual already believed to be false (e.g., “Candidate A is not a crook”) led to an increase in the perceived validity of what was denied (i.e., “Candidate A is a crook”), presumably because conversational logic (Grice, 1975) led recipients to reason that the source would not make a denial unless there was something to the charge (see also Wegner, Wenzlaff, Kerker, & Beattie, 1981). Our point here is that some validity mechanisms do not require the individual to think actively about antecedent conditions that would bring about the propositional event. Later in this chapter, we discuss both effort­ ful and more simple means of rendering evaluative judgments. Organization o f Attitude-Relevant Information and Implications for Attitude Change. Although considerable research has addressed the contents of an atti­ tude structure in memory, relatively little work has addressed the specific or­ ganization of that information. That is, the information and experiences relevant to an attitude object can be well organized or poorly organized; if well organized, a number of systems are possible. For example, consistent with the tripartite model, it is possible that attitudes are organized into affective, cognitive, and behavioral subsystems (e.g., Breckler, 1984). Another possibility is that attitude structures are bipolar with attitude-consistent information clustered together and linked to a cluster of attitude-inconsistent information (e.g., Judd & Kulik, 1980; Pratkanis, 1989). This bipolar structure could cut across affective, cognitive, and behavioral categories, or could be the organizing structure within these categories. Alternatively, the information in memory could be organized in some other fashion, such as into the categories of acceptable, objectionable, and noncom­ mittal (e.g., Sherif & Hovland, 1961).6 6Although there is considerable persuasion research in the social judgment theory tradition (Sherif, Sherif, & Nebergall, 1965), it is not clear that the measured latitudes of acceptance, rejection, and

80

PETTY, PRIESTER, WEGENER

A currently popular view of attitude structure relies on an associative network model of memory (e.g., Anderson, 1983), in which an attitude object is linked to an evaluative node as well as to other relevant information and experiences (e.g., Pratkanis & Greenwald, 1989). Although most social psychologists have focused on the structure of individual attitudes (intra-attitudinal structure; e.g., Judd & Krosnick, 1989), any one attitude structure also can be linked to other attitude structures that have some basis of similarity (e.g., derive from the same values; inter-attitudinal structure; see Eagly & Chaiken, in press). Because of “spreading activation,” the activation of one concept spreads to linked nodes in the system. For example, in one study, individuals were faster to respond to an attitude item when it followed a question on a similar topic than when it followed a dissimilar one (Tourangeau, Rasinski, & D’Andrade, 1991). In general, items in an attitude structure will be more or less accessible depending on the frequen­ cy and recency of their activation and use (Higgins, 1989; Wyer & Srull, 1986). Although understanding the organization of attitude-relevant material in memory undoubtedly awaits further developments in understanding the structure of memory in general, we nevertheless can summarize a few useful ideas about attitude structure and attitude change that are accepted by most attitude theorists. First, changing some aspect of an individual attitude structure most likely will lead to some change in the attitude, although it might take some time and thought for the change to become manifest (McGuire, 1981). That is, if attitudes are based on beliefs, emotions, and/or behaviors, and are linked in memory to them, then a strengthening or weakening of any of these links or the addition of new links potentially could modify the attitude. Second, the more extensive the structure supporting an individual attitude, the more difficult it should be for any one new piece of information to modify the attitude, especially if the structural elements are highly accessible. Third, to the extent that any one attitude is linked to many others, it can be viewed as central in a person’s attitudinal system (Rokeach, 1968) or high in embeddedness (Scott, 1968). To the extent that a central attitude changes, others are likely to change as well.7 Attitude Strength

Attitude structure is related to another important property of attitudes—attitude strength. Strong attitudes are those that are stable over time; resistant to counter­ persuasion; and likely to influence other judgments, intentions, and behaviors (Petty & Krosnick, in press). It is useful to consider attitude strength as a con­ tinuum (e.g., Fazio, 1986). At one end are non-attitudes (that is, no attitude noncommitment correspond to the organization of attitude-relevant material in memory. Furthermore, the specific attitude change predictions of this model have not received strong support (Eagly & Telaak, 1972). 7Central attitudes sometimes are considered values (e.g., one’s attitude toward the abstract con­ cept of “equality” can have considerable impact on many other more specific attitude objects).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

81

exists about the object), and at the other end are attitudes that are quite strong. Most attitudes, of course, fall somewhere in between. In general, the stronger the attitude, the more it possesses the characteristics noted earlier. For example, the stronger the attitude, the more it tends to have an influence on a person’s be­ havior. A number of measurable properties of attitudes have been related to the afore­ mentioned strength consequences. Specifically, attitudes that are polarized or ex­ treme (e.g., Abelson, in press; Fazio & Zanna, 1978; Judd & Brauer, in press), important (Krosnick, 1988; see Boninger et al., in press), or high in certainty (e.g., Davidson, Yantis, Norwood, & Montano, 1985; see Gross, Holtz, & Miller, in press; Pelham, 1991), attitude-belief consistency (e.g., Chaiken & Baldwin, 1981; Norman, 1975), direct experience (e.g., Davidson et al., 1985; Regan & Fazio, 1977), personal relevance (e.g., Howard-Pitney, Borgida, & Omoto, 1986; Petty & Cacioppo, 1986a; see Thomsen, Borgida, & Lavine, in press), vested interest (see, Crano, in press), knowledge (e.g., Davidson, in press; Wilson, Kraft, & Dunn, 1989; Wood, 1982; see Wood, Rhodes, & Biek, in press), and accessi­ bility (e.g., Bassili & Fletcher, 1991; see Fazio, in press; Fazio & Williams, 1986) all produce one or more of the strength consequences (see Krosnick & Abel­ son, 1992; Petty & Krosnick, in press). Although it would simplify matters con­ siderably if attitude strength was a unitary concept, research has suggested that these properties show relatively low correlations with each other (Krosnick, Boninger, Chuang, Berent, & Carnot, in press; Raden, 1985). Thus, there ap­ pears to be a variety of determinants of attitude strength. Attitude strength is of concern to persuasion researchers for at least two rea­ sons. First, as is seen shortly, some attitude change processes are especially ef­ fective in changing weak, rather than strong, attitudes. Second, some attitude change processes are more effective in producing strong attitudes than others. In the remainder of this chapter, we focus on the processes responsible for changes in attitudes and on the consequences of these changes. Our focus is on the the­ ories that presently generate the bulk of empirical work on attitude change (see Insko, 1967; Kiesler, Collins, & Miller, 1969; and McGuire, 1985, for excel­ lent reviews of earlier approaches).

INFORMATION PROCESSING STAGES IN ATTITUDE CHANGE

Now that we have reviewed the basic nature of attitudes and discussed their un­ derlying bases and structure, we can turn to the specific theories and processes of attitude change that have developed. We begin by discussing a general infor­ mation processing approach that has been popular in one form or another for a considerable period of time. A typical information-processing model of persu­ asion views attitude change as the culmination of a series of information-processing

82

PETTY, PRIESTER, WEGENER

stages. For example, in 1925, Strong proposed that people first needed to be ex­ posed to a message and pay attention to it before it could influence them. However, the contemporary impetus for the information-processing approach can be found in the Yale Communication Program led by Carl Hovland (Hovland, Janis, & Kelley, 1953). The Hovland group was influential in its view that attitude change resulted if there was attention, comprehension, learning, acceptance, and reten­ tion of the message and its conclusion. A person would engage in these steps to the extent that the persuasive message or the persuasion context provided in­ centives for doing so. Over the years, McGuire (1968, 1989) has presented a more formal information-processing model that incorporates and extends these ideas. Although the number of steps in McGuire’s information-processing model has varied, a consistent theme has been that persuasion was dependent on various factors related to the reception of message arguments and various factors related to yielding to them (see Wyer, 1974, for a mathematical specification and elabo­ ration of this model). Reception/Learning Processes

McGuire’s (1968) reception/yielding model holds that both reception and yield­ ing processes are required for attitude change. Thus, some variables might decrease the likelihood of persuasion by making reception more difficult (e.g., a complex or incomprehensible message), whereas others might decrease persuasion by mak­ ing yielding less likely (e.g., a poorly reasoned message). Perhaps the most in­ teresting variables in this framework were those that were postulated to influence each step differently. For example, as audience intelligence increased, reception might increase (e.g., because greater intelligence should produce greater com­ prehension and memory), but yielding might decrease (e.g., because an intelli­ gent audience has greater confidence in its initial opinions or can better resist the arguments presented; Eagly & Warren, 1976). For variables having an oppo­ site impact on reception and yielding processes, persuasion should be maximal at a moderate level of the variable (e.g., moderate intelligence). Although some studies examined the curvilinear hypothesis from this model directly (see Rhodes & Wood, 1992), most of the attention devoted to the recep­ tion/yielding model was on examining the notion that reception was a prerequi­ site to persuasion. In particular, it seems obvious that a recipient at least needs to comprehend the position taken in a message if change toward that position is to take place. Consequently, most attention focused on the more open question of whether a recipient needs to comprehend or understand the supporting infor­ mation or arguments offered in the message for persuasion to occur. Thus, researchers have examined the extent to which argument reception, as assessed via message recall or recognition, is related to persuasion. Although a measure of message recall requires more than the mere initial reception, comprehension, or learning of the message arguments, the learning theory approach to persuasion

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

83

also included the idea that attitude change persistence should be related to mes­ sage persistence. Because of this, message recall became a popular assessment technique (McGuire, 1968). However, there is surprisingly little evidence that either initial persuasion or its persistence depends on a person’s ability to learn and remember the arguments in a persuasive communication (see Cook & Flay, 1978; Eagly & Chaiken, 1993; McGuire, 1969, for reviews).8 There are several reasons for low attitude change-argument memory rela­ tionships. Perhaps the most important is that a pure reception model fails to consider the idiosyncratic evaluations individuals have of the arguments. That is, one person can find an argument very compelling, another only slightly convincing, and a third might find the same argument to be completely ridicu­ lous (Petty, Ostrom, & Brock, 1981). As McGuire (1968) postulated, persua­ sion should depend on yielding to the arguments in addition to reception of them. When recalled information is weighed by a person’s idiosyncratic evalua­ tion of it, attitude-recall relationships are increased (e.g., Chattopadhyay & Alba, 1988). Second, argument learning and recall might not relate to persuasion even if perceptions of argument quality were controlled. This is because a person might not find all of the arguments equally relevant to assessing the validity of the mes­ sage or equally diagnostic in allowing a comparison among alternatives (Feld­ man & Lynch, 1988). If a person is thinking about a message as it is received and is evaluating the information it contains, it is likely that an attitude will emerge following the communication. To the extent that both relevant and irrelevant ar­ guments are equally accessible immediately following message exposure, but only relevant arguments enter into the attitude (e.g., arguments that match the func­ tional basis of an attitude), an inconsistency between recall and persuasion would occur. On the other hand, suppose a person is not forming an attitude as a message is received, but must express an opinion at a later point in time. Then, the judg­ ment is more likely to depend on the implications of the information that can be recalled if that information is perceived relevant at the time of judgment. Researchers who have distinguished between attitudes formed during informa­ tion exposure (spontaneous or on-line judgments) versus judgments formed sub­ sequent to message exposure have shown that the latter type of judgment is related more strongly to the information recalled than is the former (Bargh & Thein, 1985; Hastie & Park, 1986; Lichtenstein & Srull, 1985). 8Of course, if the message appears to be gibberish or odd, persuasion will be reduced (Eagly, 1974), but accurate comprehension or learning may not be necessary. In fact, misunderstanding even could lead to more persuasion than correct perceptions. That is, an individual’s subjective perception of understanding is probably more important than actual understanding. Some theorists have argued that the importance of reception and learning in the typical persuasion study has been underestimat­ ed, because the variance in reception in a typical laboratory experiment may be too restricted to de­ tect significant effects (Eagly & Chaiken, 1984).

84

PETTY, PRIESTER, WEGENER

Thus, factors that increase a person’s motivation and ability to think about and elaborate information at the time of message exposure should reduce the correla­ tion between attitudes and information recalled from the message. Tesser and Shaffer (1990) speculated that two-sided messages on unfamiliar topics would reduce the likelihood of elaboration when compared with one-sided communica­ tions, and therefore should increase the relative importance of message learning. Consistent with this reasoning, one study found that when recipients were given two sets of persuasive arguments, both message recall and cognitive elaborations predicted persuasion; but when only one set of arguments was received, elabora­ tions predicted attitudes but recall did not (Chattopadhyay & Alba, 1988). In another study (Mackie & Asuncion, 1990), individuals who were distracted from elaborating a persuasive message showed higher memory-attitude relationships than individuals who received the message under high elaboration conditions. Finally, in a study examining individuals who differed in their motivation to think about a message (Haugtvedt & Petty, 1992), low motivation subjects showed higher memory-attitude correlations than those who were more motivated to think. Because the messages presented in these studies contained reasonably strong arguments, it is not clear if the memory-recall relationships that were ob­ served were due to the fact that people favorably evaluated the arguments they recalled at the time of judgment, or whether they simply made inferences of va­ lidity based on the mere number of arguments that they could recall (Petty & Cacioppo, 1984a). The research outlined earlier suggested some circumstances in which message recall would be related to post-message attitudes. Specifically, the correlation between message recall and attitudes should be increased if: (a) the likelihood of elaborating the message arguments at the time of message exposure is rela­ tively low, (b) an unexpected judgment is required following message exposure, and (c) no simple cues (e.g., source expertise) are available or attended to by the recipient that could produce an attitude at the time of judgment in the absence of on-line argument scrutiny. In sum, after a period in which persuasion theorists generally dismissed re­ call and learning processes from the list of persuasion mediators, current re­ search has documented conclusively that there is a specific role for a memorybased approach. In addition to work on the conditions under which argument recall will be related to attitudes, current work on learning and memory pro­ cesses in persuasion is focused on questions such as: (a) the situational and contextual determinants of what information will be recalled (e.g., Kisielius & Sternthal, 1984, 1986; MacKenzie, 1986; Schul & Burnstein, 1983; Sherer & Rogers, 1984; Weldon & Malpass, 1981); (b) how arguments are represented in memory (e.g., Schmidt & Sherman, 1984; Wilder, 1990); and (c) individual differences in message recall that are either chronic or issue specific (e.g., Cacioppo, Petty, & Morris, 1983; Chaiken et al., in press; Reardon & Rosen, 1984).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

85

Evaluation/Yielding Processes

As our discussion implies, tracking the mere reception of information is insuffi­ cient to gauge the persuasiveness of a communication. Rather, the processes by which a person reacts to the incoming stimuli must be understood. Our discus­ sion of the processes by which people yield to persuasive communications is divid­ ed into three parts. First, we discuss the general cognitive response idea that it is important to understand a person’s idiosyncratic reactions to incoming infor­ mation or a person’s self-generation of information instead of (or in addition to) their reception and learning of the information presented (Greenwald, 1968; Pet­ ty et al., 1981). Next, we discuss the notion inherent in the elaboration likeli­ hood model of persuasion (Petty & Cacioppo, 1981, 1986b) that a person’s idi­ osyncratic reactions can be based on considerable cognitive effort or on more simplistic analyses (see also the heuristic/systematic model; Chaiken, Liberman, & Eagly, 1989). Finally, we discuss various yielding processes that do not rely on careful and effortful scrutiny of the information contained in a persuasive ap­ peal. Cognitive Response Approach

Cognitive Responses to Messages

The cognitive response approach to persuasion was developed as an explicit attempt to salvage a learning view of persuasion when it was challenged by numer­ ous findings of low attitude-recall correlations. Greenwald (1968) proposed that it was not the specific arguments in a message that were paired with the message conclusion (or attitude object) in memory as the Yale approach implied, but rather a person’s idiosyncratic cognitive responses or reactions to the message arguments that were paired with the conclusion and thus responsible for persuasion (see also, Brock, 1967). That is, the cognitive responses were postulated to be linked to the global evaluation of the attitude object, but the initial information on which the responses were based were stored elsewhere in memory (Anderson & Hubert, 1963). In addition to this structural difference, the cognitive response approach provided an alternative to message reception and learning as yielding mechanisms. Of course, at the most basic level, individuals only can respond cognitively to something that they have received from the message. However, in the cognitive response analysis, a person’s thoughts can concern incorrectly perceived argu­ ments as well as correctly perceived arguments. In fact, the thoughts can be to the message conclusion in the absence of receipt of any arguments. Further­ more, the cognitive response approach was sufficiently general that it accom­ modated cognitive responses elicited by the tone of the message or by extra message factors, such as the source of the message or the context in which it was presented. In any case, to the extent that a person’s cognitive responses were favorable, persuasion was the postulated result; but to the extent that they were

86

PETTY, PRIESTER; WEGENER

unfavorable (e.g., counterarguments, source derogations), resistance or even boomerang were possible (Petty et al., 1981). Greenwald (1968) further proposed that persistence of persuasion depended on the decay function for cognitive responses, rather than the message arguments per se (Love & Greenwald, 1978). The cognitive response approach generated a considerable body of evidence consistent with the view that, in certain situations, people spontaneously produce thoughts during message presentation, and the favorability of these thoughts is a good predictor of post-message attitudes and beliefs (Eagly & Chaiken, 1984, 1993; Perloff & Brock, 1980; Petty & Cacioppo, 1986a). In a typical cognitive response study, message recipients list or verbally report their thoughts either during or after the message. Studies in support of the cognitive response approach showed that: (a) physiological activity indicative of information processing (e.g., speech EMG) is elevated when cognitive responding is presumed to occur (e.g., Cacioppo & Petty, 1979a); (b) thought profiles show the same pattern as the atti­ tude measure in response to some manipulation (e.g., the manipulation produces increased persuasion and increased favorable thoughts and/or decreased unfavor­ able thoughts; e.g., Osterhouse & Brock, 1970); (c) the polarity of these thoughts (e.g., positive minus negative thoughts) is a good predictor of the post-message attitude (e.g., Mackie, 1987); and (d) removing the effect of some manipulation on thoughts eliminates its effect on attitudes, whereas the reverse does not occur (e.g., Insko, Turnbull, & Yandell, 1974; see Cacioppo & Petty, 1981b; Wright, 1980; for reviews). Although most studies simply have categorized thoughts into favorable, un­ favorable, and neutral categories, other coding schemes are possible. For exam­ ple, thoughts that implicate the self are better predictors of attitudes than thoughts that do not (Shavitt & Brock, 1986). It is also possible to distinguish between cognitive responses that are based on argument scrutiny (e.g., counterarguments) versus thoughts that are based on simple cues (e.g., source derogations). As the more recent theories of persuasion that we discuss in the next section made ap­ parent, either type of thought can be a better predictor of attitudes under certain circumstances (e.g., Chaiken, 1980; Petty & Cacioppo, 1979b, 1984a; Wright, 1974).9 Cognitive Responses in the Absence of a Message

Self-Generation o f Arguments. Just as a person’s thoughts in response to a persuasive message can determine the extent and direction of attitude change, so too can a person’s thoughts in the absence of any explicit external message. The powerful and persisting effects of completely self-generated messages was 9Additional cognitive response coding schemes have been reported (e.g., recipient generated or externally originated thoughts, Greenwald, 1968; cognitive versus affective reactions, Batra & Ray, 1986; Ickes, Robertson, Tooke, & Teng, 1986), but there are too few studies to assess their utility.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

87

shown in early research on role-playing (Janis & King, 1954; Watts, 1967). In this research, people typically were asked to act out certain roles (e.g., convince a friend to stop smoking) or generate a message on a certain topic, and the subse­ quent attitudes of these people were compared with those of controls who had either passively witnessed the role-playing, or listened to a communication, or received no message. A consistent research result was that active participation in the generation of a message is a successful strategy for producing attitude change (e.g., McGuire & McGuire, 1992), and that these changes tend to persist longer than changes based on passive exposure to a communication (e.g., Elms, 1966; Huesmann, Eron, Klein, Brice, & Fischer, 1983). In addition, some research has indicated that people find their own arguments to be more original than those that are gener­ ated by others, and self-generated arguments are also more memorable (Green­ wald & Albert, 1968; Slamecka & Graf, 1978). Presentation of arguments in public, rather than in private, enhances attitude change due to role-playing, perhaps because people try harder in public, the arguments are more salient, or counterarguing of the role-playing is more likely in private (Tice, 1992). Generating Explanations and Imagining Events. In another stream of experi­ ments, the effects of asking people to generate explanations for some proposition or to imagine the occurrence of some event were examined. For example, in one study, students were presented with detailed case histories that led them to ex­ plain why being a good fireman was associated with either a high or a low will­ ingness to take risks (Anderson, Lepper, & Ross, 1980). Later, it became clear that the case histories on which the explanations were based were completely false. Nevertheless, subjects who were led to think about why a high willingness to take risks predicted success in fire fighting continued to believe in this relation­ ship, whereas those led to think about the opposite continued to believe the op­ posite. Similarly, people who were asked to imagine hypothetical events (e.g., that Ohio State would beat UCLA in the Rose Bowl) came to believe that these events have a higher likelihood of occurring than before thinking about them (e.g., Sher­ man, Cialdini, Schwartzman, & Reynolds, 1985). Consistent with the earlier work on role-playing, the work on generating explanations and imagining events showed consistently that self-generation is a powerful way to change beliefs, and that these beliefs are remarkably impervious to change (see also Lepper, Ross, & Lau, 1986; Sherman, Zehner, Johnson, & Hirt, 1983). Research on Mere Thought. Finally, in an extensive series of studies, Tes­ ser (1978) and his colleagues examined the effects of merely asking someone to think about an issue, object, or person (see Tesser, Martin, & Mendolia, in press, for a review). For example, in one early study, Sadler and Tesser (1973) in­ troduced subjects to a likable or dislikable partner (via a tape recording). Some

88

PETTY, PRIESTER, WEGENER

of the subjects were instructed to think about the partner, whereas others were distracted from doing so. The thinking manipulation polarized judgments of the partner. Specifically, enhanced thinking was associated with more favorable evalu­ ations of the likable partner, but less favorable ratings of the dislikable partner. Current research has indicated that both moderation and polarization can result from mere thought. Specifically, the polarization effect requires that subjects have a well-integrated and consistent schema to guide processing and that they are moti­ vated to utilize this issue-relevant knowledge (Chaiken & Yates, 1985; Liberman & Chaiken, 1991; Tesser & Leone, 1977). In the absence of these conditions, such as when motivation to think is low or when the issue-relevant information in memory represents independent dimensions of knowledge rather than a highly interconnected (correlated) system of beliefs, mere thought is associated with at­ titude moderation (e.g., Judd & Lusk, 1984; Linville, 1982; Millar & Tesser, 1986a). The Elaboration Likelihood Model of Persuasion

The cognitive response approach and research on self-generated attitude and be­ lief change demonstrated quite conclusively that active thought processes often accompany attitude change, and that self-generated change can be quite endur­ ing. Although the cognitive response approach in its broadest framework appears to provide a reasonable account of initial attitude change and its persistence, its focus on the very active cognitive evaluation of the message (or self-generation of arguments) implies that little attitude change is likely when active thinking is low. Yet, a number of persuasion investigations have suggested that it actually might be easier to change attitudes when people have relatively little interest in or knowledge about the topics of the persuasive message—conditions that would not be likely to foster much issue-relevant cognitive activity (see Hovland, 1959; Johnson & Eagly, 1989; Petty & Cacioppo, 1986a). Thus, active thinking about the substance of a message does not seem to be a requirement for attitude change.

Central and Peripheral Routes to Persuasion

The elaboration likelihood model of persuasion (ELM; Petty & Cacioppo, 1981, 1986b) is a theory about the processes responsible for yielding to a persuasive communication and the strength of the attitudes that result from those processes. The model holds that the processes responsible for yielding to a communication can be thought of as emphasizing one of two relatively distinct routes to persua­ sion. The first, or central route, involves effortful cognitive activity, whereby the person draws on prior experience and knowledge to assess and evaluate care­ fully the central merits of the position put forth in the persuasive message. Con­ sistent with the cognitive response approach to persuasion, the message recipient

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

89

under the central route is actively scrutinizing all of the information presented. The goal of this cognitive effort is to determine if the position advocated by the source has any merit. The end result of the effortful information processing in­ volved in the central route is typically an attitude that is well articulated and bol­ stered by supporting information. In contrast, attitudes also can be changed by a peripheral route without much thinking about information central to the merits of the attitude issue. The peripheral route recognizes that it is neither adaptive nor possible for people to exert con­ siderable mental effort in thinking about all of the messages and attitude objects to which they are exposed. To function in life, people sometimes must act as “lazy organisms” (McGuire, 1969) or “cognitive misers” (Taylor, 1981) and em­ ploy simpler means of evaluation. That is, in some situations, considerable cog­ nitive resources are allocated to the task of processing the message, but at other times few resources are delegated. The peripheral route characterizes attitude change as resulting from relatively low resource demanding processes, such as when rewards, punishments, and affective experiences are associated directly with an attitude object (as in classical or operant conditioning; e.g., Staats & Staats, 1958); or when generated inferences (as in the self-perception mandate “I bought it, so I must like it”; Bern, 1972), well-learned heuristics (as in “Experts are gener­ ally correct”; Chaiken, 1980, 1987; Cialdini, 1987), or category-based process­ ing (“She’s a liberal, so I don’t like her”; Fiske & Pavelchak, 1986) determine attitudes. Attitudes that are changed by these peripheral route processes are postu­ lated to be weaker generally than attitudes that are changed by the central route.10 The central and peripheral routes to persuasion anchor an elaboration likeli­ hood continuum. When situational and dispositional factors render the likelihood of message elaboration high, the central route process should dominate; but as the elaboration likelihood decreases, peripheral route processes should become more evident. Persuasion researchers have identified a number of ways to assess the extent of information processing activity. Perhaps the most popular proce­ dure has been to vary the quality of the arguments contained in a message and to gauge the extent of message processing by the size of the argument quality 10One of the earliest theories that explicitly acknowledged different processes underlying atti­ tude expression was Kelman’s (1958) three-process model. Kelman tied the type of persuasion large­ ly to the source of the message—experts produced internalization (a permanent type of change based on personal acceptance of the message conclusion as one’s own); attractive sources produced identifi­ cation (change that was tied to a continued association with a desirable referent); and powerful sources produced compliance (change that was expressed only under the continued threat of rewards and punish­ ments controlled by the source). Kelman’s distinction of the different types of change was useful, even though, as is explained in this chapter, the determinants of those changes have proved to be much more complicated than he suspected initially. The term internalization can be useful when referring to changes induced by the central route. However, the term identification refers to only one process by which attitudes can be changed by the peripheral route. The term compliance does not refer to real change at all, but only that expressed when the person is under appropriate scrutiny.

90

PETTY, PRIESTER, WEGENER

effect on attitudes (Petty, Wells, & Brock, 1976; see Johnson & Turco, 1993, for a review). Greater argument quality effect sizes, of course, suggest greater argument scrutiny. In addition to the argument quality procedure, several other methods for as­ sessing message processing have been used. For example, investigators have ex­ amined the number and profile of issue relevant thoughts generated (Petty et al., 1981). High elaboration conditions are associated with more thoughts (e.g., Burnkrant & Howard, 1984) and thoughts that better reflect the quality of the argu­ ments presented (e.g., Harkins & Petty, 1981). Also, correlations between message-relevant thoughts and post-message attitudes tend to be greater when argument scrutiny is high (e.g., Chaiken, 1980; Petty & Cacioppo, 1979b), and high message elaboration produces longer reading or exposure times than more cursory analyses (Mackie & Worth, 1989). Our discussion so far has highlighted two ways in which variables can have an impact on persuasion. Sometimes variables are relevant to the central merits of an object, and therefore can serve as persuasive arguments. Also, variables can serve as peripheral cues, allowing favorable or unfavorable attitude forma­ tion in the absence of a diligent consideration of the central merits of the object or issue. In addition, variables can have an impact on persuasion by influencing the extent of elaboration (i.e., the degree to which the person is motivated and/or able to evaluate the merits of the issue-relevant information presented). Finally, variables can influence the direction of thinking (i.e., whether the thoughts elicited are relatively favorable or unfavorable). Each of these roles for variables is most likely under a specific set of conditions, as we describe shortly. Motivation Versus Ability

The ELM holds that there are many variables capable of affecting the elabo­ ration likelihood and thereby influencing the impact of the processes falling un­ der each route to persuasion. Some variables affect a person’s motivation to process issue-relevant arguments, whereas others affect ability or opportunity. Some vari­ ables are part of the persuasion situation, whereas others are part of the individu­ al. Some variables affect mostly the extent of information processing activity, whereas others tend to influence the direction of the thinking. Table 2.1 illus­ trates variables falling into each cell of the 2 x 2 x 2 matrix. Motivation to Think. Perhaps the most important variable influencing a per­ son’s motivation to think is the perceived personal relevance of the communica­ tion. For example, in one study, Yale undergraduates generated significantly more thoughts when they were asked to think about personally relevant events than about equally desirable events that were higher in societal importance but lower in personal significance (McGuire & McGuire, 1991). Personal or self-relevance can stem from a variety of sources, such as the attitude object being linked to

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

91

TABLE 2.1 Motivational and Ability Factors Influencing the Extent and Direction of Processing M otivational Factors Variable

Situational

Ability Factors

Individual

Situational

Individual

Extent of Processing

Induced personal relevance

Need for cognition

External distraction

General intel­ ligence

Direction of Processing

Forewarning of in­ tent to persuade

Open/closed mindedness

Instructed head movements

Attitude-congruent knowledge

personal values, outcomes, groups, possessions, or individuals that are impor­ tant to the individual (Johnson & Eagly, 1989; Petty, Cacioppo, & Haugtvedt, 1992; Thomsen et al., in press). When the personal importance of an issue is high, people are motivated to scrutinize the information in a message, and atti­ tude change is based largely on the quality of the information presented in sup­ port of the issue (Leippe & Elkin, 1987; Petty & Cacioppo, 1979b, 1990). Merely changing the pronouns in a message to enhance self-relevance (e.g., you vs. peo­ ple) can increase information processing activity (Burnkrant & Unnava, 1989), as can enhancing self-awareness by placing message recipients in front of a mir­ ror (Hutton & Baumeister, 1992). When personal relevance is low, argument scrutiny is reduced and attitudes are affected more by variables serving as peripher­ al cues, such as the status, likability, or attractiveness of the message endorsers (Chaiken, 1980; Miniard, Bhatla, Lord, Dickson, & Unnava, 1991; Petty, Caciop­ po, & Schumann, 1983; Schumann, Petty, & Clemons, 1990).11 Of course, variables other than personal relevance can modify people’s moti­ vation to think about a message. For example, people are more motivated to scrutinize information when they believe that they are solely responsible for mes­ sage evaluation (Petty, Harkins, & Williams, 1980). People also process more when they are individually accountable (Tetlock, 1983), and when they recently have been deprived of control (Pittman, in press). Increasing the number of mes­ sage sources can enhance information processing activity (Harkins & Petty, 1981; Moore & Reardon, 1987), especially when the sources are viewed as providing independent assessments of the issue (Harkins & Petty, 1987). On the other hand, 11Current work on personal relevance has stemmed from previous work in social psychology on ego-involvement conducted within the context of social judgment theory (Sherif & Sherif, 1967). Many specific definitions and models of involvement have appeared in the literature. Some defini­ tions and models of involvement are quite narrow, but others encompass nearly every factor that can affect any type of information processing activity (cf. Greenwald & Leavitt, 1984). In this chap­ ter, we talk about specific variables (e.g., personal relevance, knowledge, etc.) that affect a person’s motivation or ability to process the merits of an attitude object, rather than dwell on the many defini­ tions and meanings of the involvement concept.

92

PETTY, PRIESTER, WEGENER

messages that are overly quantitative increase reliance on peripheral strategies, because people are less motivated or able to process them (Yalch & Elmore-Yalch, 1984). Messages that are moderately inconsistent with an existing attitude sche­ ma can enhance processing over schema-consistent messages, because the for­ mer could pose some threat that needs to be understood or some incongruity that needs to be resolved (Cacioppo & Petty, 1979b; Mandler, 1982; Meyers-Levy & Tybout, 1989). Other incongruities also increase information processing ac­ tivity, such as when an expert source presents surprisingly weak arguments (Maheshwaran & Chaiken, 1991). In addition to factors associated with the persuasive message or the persuasion context, there are individual differences in people’s motivation to think about per­ suasive communications. For example, people who enjoy thinking (i.e., those high in need for cognition; Cacioppo & Petty, 1982) tend to form attitudes on the basis of the quality of the arguments in a message, thereby following the central route to persuasion (Cacioppo et al., 1983). People who do not enjoy thinking are more reliant on simple peripheral cues in the persuasion context (Axsom, Yates, & Chaiken, 1987; Cacioppo & Petty, 1984; Haugtvedt, Petty, & Cacioppo, 1992). Recent research has suggested that because high need for cognition individu­ als think intently about messages, this can produce primacy effects in judgment when they are presented with opposing communications from different sources. In particular, a primacy effect results when extensive processing of an initial mes­ sage allows people high in need for cognition to resist or counterargue the subse­ quent message (Haugtvedt & Wegener, 1993; see also, Haugtvedt & Petty, 1992). Primacy could also be produced if thinking about the first message is so intense that it disrupts or interferes with thinking about a second communication that im­ mediately follows it (Bozzolo, 1993). In contrast, individuals low in need for cognition are less likely to process either message and may therefore show recency effects due to greater utilization of information from the second message (Haugt­ vedt & Wegener, 1993). Active processing of the initial incoming information may explain why people high in need for cognition interpret ambiguous evidence from mock trial testimony as consistent with introductory comments that frame the evidence as either pro-defense or pro-prosecution (a primacy effect). In con­ trast, people low in need for cognition interpret the same evidence as consistent with concluding comments that frame the evidence as supporting the side opposed to the introductory comments (a recency effect that is perhaps due to memory processes; Kassin, Reddy, & Tulloch, 1990). On the other hand, in some judg­ ment settings, low effort processing might lead to greater rather than reduced primacy. For example, if the two sides of an issue are presented by one source or in one communication rather than in two, low effort information processors might be more likely than highly motivated processors to form a snap judgment prior to receipt of all of the information (Ahlering & Parker, 1989).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

93

Ability to Think. Among the important variables influencing a person’s abili­ ty to process issue-relevant arguments is message repetition. Moderate message repetition provides more opportunities for argument scrutiny (e.g., Cacioppo & Petty, 1979b; Gorn & Goldberg, 1980; Rethans, Swasy, & Marks, 1986), which will prove beneficial for persuasion as long as the arguments are strong and tedi­ um is not induced (Batra & Ray, 1986; Cacioppo & Petty, 1989; Cox & Cox, 1988). External distraction (e.g., Petty et al., 1976) and a fast presentation (Smith & Shaffer, 1991) reduce argument elaboration, which can be beneficial when the message would have been counterargued easily. People are also generally better able to process messages that appear in print than messages that are con­ trolled externally (e.g., radio and TV; Chaiken & Eagly, 1976; Wright, 1980). On the other hand, placing time pressures on processing (Kruglanski & Freund, 1983), enhancing physiological arousal via exercise (Sanbonmatsu & Kardes, 1988), or rendering the message difficult to understand (Ratneshwar & Chaiken, 1991) increases reliance on simple cues. There are also individual differences in the ability of people to think about a persuasive communication. For example, as general knowledge about a topic increases, people become more able (and perhaps more motivated) to think about issue-relevant information (Wood et al., in press). However, knowledge is only effective to the extent that it is accessible (Brucks, Armstrong, & Goldberg, 1988). When knowledge is low or inaccessible, people are more reliant on simple cues (Burke, DeSarbo, Oliver, & Robertson, 1988; Wood & Kallgren, 1988; cf. Alba & Hutchinson, 1987). For example, increasing the mere amount of information associated with one side over another influences attitudes more for low than for high knowledge individuals (Alba & Marmorstein, 1987; Wood, Kallgren, & Preisler, 1985). Combining Variables. In most communication settings, a confluence of fac­ tors determines the nature of information processing, rather than one variable acting in isolation. For example, the use of rhetorical questions can increase think­ ing about a persuasive message if the questions follow the arguments and moti­ vation to think about the message would normally be low (Petty, Cacioppo, & Heesacker, 1981). On the other hand, if people already are motivated to think about the message or the rhetorical questions precede the message arguments, then the use of rhetoricals actually can disrupt the normal processing that would have occurred (Howard, 1990; Petty et al., 1981; Swasy & Munch, 1985). As another example, message repetition is particularly important as a means of in­ creasing thinking when the message is relatively difficult to process, but not when it is easy (Batra & Ray, 1986). The next decade is likely to see more research on the interacting effects of variables affecting motivation and ability to elaborate messages.

94

PETTY, PRIESTER, WEGENER

Relatively Objective Versus Biased Information Processing

The variables we discussed earlier, such as distraction or need for cognition, tend to influence information processing activity in a relatively objective man­ ner. That is, all else being equal, distraction tends to disrupt whatever thoughts a person is thinking (Petty et al., 1976). The distraction does not specifically tar­ get one type of thought (e.g., favorable or unfavorable) to impede. Similarly, individuals with high need for cognition are more motivated to think in general than people low in need for cognition (Cacioppo et al., 1983). They are not more motivated chronically to think certain kinds of thoughts over others. Biases can emerge in the processing of a persuasive message in a number of ways. First, some people simply might possess a biased store of knowledge com­ pared with other people. If so, their ability to process the message objectively can be compromised. Second, variables in the persuasion situation can bias retrieval of information even if what is stored is completely balanced. For exam­ ple, a positive mood (Bower, 1981) or vertical head movements (Tom, Pettersen, Lau, Burton, & Cook, 1991; Wells & Petty, 1980) might bias retrieval in favor of positive over negative thoughts. People also can be biased in their active search for information in memory that would support one point of view over another (Petty & Cacioppo, 1977). For example, when people are asked to present information to an audience whose opinions are known, their communications to these audiences tend to be biased in favor of the audience’s opinions (e.g., Hig­ gins & McCann, 1984; McCann, Higgins, & Fondacaro, 1991). Not surprising­ ly, given the work on self-generated messages that we reviewed earlier, people’s own subsequent attitudes are biased by the biased communications they generate (see McCann & Higgins, 1984, for a review). Finally, biases can have an impact on the integration of information. That is, even if people have an unbiased store of information in memory, and a balanced set of ideas is retrieved, certain ideas might be given greater weight than others in forming a judgment (Anderson, 1981).12 Ability Factors That Produce Bias. Although many theorists have focused on how motivational factors and a person’s goals can bias information processing activity (e.g., Johnson & Eagly, 1989; Kunda, 1990), biased attitudinal outcomes can result from ability factors as well. For example, if a person can recall the message arguments that favor his or her side over the competing side (see Roberts, 1985), this does not necessarily mean that a motivated selectivity or defensive­ ness is operating. Rather, the arguments on one side of the issue might fit the existing attitude structure more readily than the opposing arguments, thereby producing a biased result. Similarly, people’s evaluations of the evidence on two 12Biases also can influence other steps in the information processing sequence. For example, peo­ ple can be biased in exposing themselves to information (Frey, 1986). However, persuasion researchers have focused on biases in processing that occur once a message has been received.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

95

sides of an issue can be influenced by their attitudes (see Houston & Fazio, 1989; Lord, Ross, & Lepper, 1979; Miller, McHoskey, Bane, & Dowd, 1993), not only because of a motivated defensiveness, but also because people’s existing knowledge structures enable them to see the flaws in the opposition arguments and elaborate the merits of their own side. Because of these factors, biased out­ comes can stem from the rather objective utilization of a biased knowledge struc­ ture in addition to a motivation to retrieve or defend ideas on one side of an issue over another. Motivational Factors That Produce Bias. A number of general motivations have been linked to biased information processing. Many of these biasing mo­ tives stem from some threat to the self such as to self-esteem, social status, or people’s freedom to hold certain beliefs and attitudes (i.e., reactance; Brehm, 1966). For example, when people are forewarned of a speaker’s explicit intent to change their attitudes on a self-relevant topic, they become motivated to counterargue the message, rather than process it objectively (Petty & Cacioppo, 1979a; see also Liberman & Chaiken, 1992). Simply being forewarned that a counterattitudinal message on a high relevance topic will be presented can increase resistance, compared with an unwarned group (Chen, Reardon, Rea, & Moore, 1992; Petty & Cacioppo, 1977). When people have a prior public commitment to an attitude position, increasing accountability leads to justification of their at­ titudes, whereas accountability in the absence of a prior commitment leads to more objective processing (Tetlock, Skitka, & Boettger, 1989). Bias in informa­ tion processing activity is facilitated when the persuasive message contains in­ formation that is ambiguous or mixed rather than clearly strong or weak (Chaiken & Maheswaran, in press; Petty, Schumann, Richman, & Strathman, 1993; Tes­ ser & Cowan, 1975). Perhaps the most studied motive that is postulated to bias information processing activity is the motive to maintain consistency (i.e., consistency among attitudes, beliefs, emotions, and behaviors; Festinger, 1957; Heider, 1958; Kiesler, 1971; Rosenberg, 1960). For example, dissonance theory holds that attitudes often can be in the service of maintaining a need for consistency among the elements in a cognitive system (Festinger, 1957, 1964). In Festinger’s original formulation of the theory, two elements in a cognitive system (e.g., a belief and an attitude, an attitude and a behavior) were said to be consonant if one followed from the other and dissonant if one implied the opposite of the other (e.g., I voted for Candidate X; he was convicted of tax evasion). Of course, two elements also could be irrelevant to each other. Festinger proposed that the psychological state of dissonance was aversive, and that people would be motivated to reduce it. One of the more interesting dissonance situations occurs when a person’s behavior is brought into conflict with his or her attitudes or beliefs (e.g., by inducing a person to write a counterattitudinal essay; Zanna & Cooper, 1974). Because behavior is usually difficult to undo, dissonance can be reduced by changing

96

PETTY, PRIESTER, WEGENER

beliefs and attitudes to bring them into line with the behavior (Festinger & Carlsmith, 1959). Although it is now clear that many of the situations described by Festinger as inducing dissonance produce physiological changes in people and an “unpleasant tension” (Elkin & Leippe, 1986; Losch & Cacioppo, 1990; see Fazio & Cooper, 1983), current research has begun to focus more on understanding the precise cause of that tension. For example, one point of view with considerable support­ ing evidence is that dissonance occurs when people believe that they are personally responsible for bringing about some foreseeable negative consequence or out­ come (e.g., Cooper & Fazio, 1984; Scher & Cooper, 1989). Thus, if a freely chosen act is inconsistent with one’s attitude, but does not result in a negative outcome, no dissonance should be produced (e.g., Cooper & Worchel, 1970). However, what is it about being responsible for negative events that people find so aversive? There are at least two points of view on this. The “self-consistency” viewpoint (Aronson, 1969; Thibodeau & Aronson, 1992) contends that if people think of themselves as competent and caring individuals, it is inconsistent with this self-view to bring about negative outcomes for oneself or others. Thus, some inconsistency between a person’s actions and his or her self-concept is at the root of dissonance. Another point of view is that it violates one’s “self-integrity” to see oneself as responsible for negative outcomes (Steele, 1988). If one assumes that nearly all people aspire to have positive self-concepts, then being responsi­ ble for negative actions produces an inconsistency between one’s actual behavior and one’s desired self-image. The self-consistency and self-integrity points of view agree that if people are given an opportunity to restore or enhance their self-esteem in some manner (e.g., encouraging them to consider ways in which they are better than other people; Tesser & Cornell, 1991; receiving validating social support for their actions; Stroebe & Diehl, 1988), dissonance reducing attitude change is less likely. The self-integrity point of view in particular focuses on the fact that virtually any bol­ stering of self-integrity can undermine dissonance effects. The self-consistency and self-integrity points of view differ in their predictions of whether high or low self-esteem people should be more susceptible to dissonance effects. The selfconsistency point of view contends that high self-esteem individuals would ex­ perience the most dissonance because producing unwanted or undesirable out­ comes is most inconsistent with the favorable self-conceptions of this group (Aronson, 1992). The self-integrity point of view (Steele & Spencer, 1993) ar­ gues that low self-esteem individuals should show stronger dissonance effects be­ cause high self-esteem individuals can more easily restore self-integrity by recruiting self images capable of affirming self-adequacy. Research evidence on this question is mixed in that some research with situationally induced self-esteem has supported the self-consistency point of view (Gerard, Blevans, & Malcolm, 1964), but other research examining chronic self-esteem has supported the self­ integrity view (Steele & Spencer, 1993).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

97

It is interesting to note that Festinger originally hypothesized that inconsisten­ cy among personally important elements (such as between one’s behavior and one’s actual or desired self-concept) would induce more dissonance than incon­ sistency among more trivial elements. Thus, some have speculated that Festinger was correct in asserting that inconsistency per se is dissonance arousing, but that many (e.g., non self-relevant) inconsistencies produce trivial (undetectable) amounts of dissonance (e.g., Berkowitz & Devine, 1989; Greenwald & Ronis, 1978). Low levels of inconsistency might produce degrees of dissonance below the “dissonance threshold” (i.e., a level of dissonance below which it is tolerated and not reduced; Wyer, 1974). In any case, it is now quite clear that various counterattitudinal actions can lead to increased cognitive activity that results in attitude change. Dissonance can result in a reanalysis of the reasons why a person engaged in a certain be­ havior or made a certain choice, and cause a person to rethink the merits of an attitude object. The end result of this reassessment can be a change in attitude toward the object.13 Corrections for Bias. Although ability and motivation-based biases in process­ ing often lead to biased persuasion outcomes, this should not always be the case. For instance, if message recipients were made aware of an event that was likely to unduly bias their perceptions, those recipients might “correct” their views in light of the biasing event (e.g., Martin, 1986; Strack, Schwarz, Bless, Kiibler, & Wanke, 1993). Such corrections might proceed in different directions depend­ ing on recipients’ perceptions of how the biasing event or stimulus (e.g., an at­ tractive source) was likely to have influenced their views (Petty & Wegener, 1993). In general, in order for corrections to occur: (a) people should have a naive the­ ory about the direction of the bias induced by some information (e.g., “I’m bi­ ased in favor of attractive people”), (b) the bias must become salient, and (c) the person must be motivated and able to make the correction. In some cases, in­ tegrative processing of the information (e.g., Schul & Burnstein, 1985) could make it difficult for people to correct an individual piece of information that con­ tributed to an overall impression. Rather, the attitude might be adjusted directly for the perceived direction of bias. 13In contrast to dissonance theory, balance theory (Heider, 1958) holds that inconsistency pres­ sures sometimes can lead to attitude change by a simple inference process, rather than because of a reanalysis of the merits of the attitude object. In particular, the theory holds that balance occurs when people agree with people they like, or disagree with people whom they dislike. This theory can account for why a person would come to like a candidate more after he or she is en­ dorsed by a favored celebrity. The theory holds that imbalance (e.g., disagreeing with someone you like) leads to attitude change toward the candidate (or the endorser) in the direction of balance (see Insko 1981, 1984, for an extended discussion). A related formulation, congruity theory, holds that attitudes toward both source and object change to restore congruity (Osgood & Tannenbaum, 1955).

98

PETTY, PRIESTER, WEGENER

Multiple Roles for Variables in Persuasion Settings

We noted that variables can serve a number of roles in persuasion settings. That is, a variable can serve as an argument or a cue, or it can affect the extent (amount) or direction (bias) of information processing. One of the powerful, al­ beit complex, features of the ELM is the recognition that any one variable can serve in multiple roles. The fact that variables can have the same impact on judg­ ments via different processes makes it essential to identify the conditions under which each process operates. Research on this problem is in its early stages, but existing literature already suggests the conditions under which variables serve in different roles (Petty & Cacioppo, 1984b). Under conditions of relatively low-elaboration likelihood, persuasion-relevant variables (e.g., source attractiveness), to the extent that they have any impact at all, serve as peripheral cues, influencing attitudes regardless of whether a mes­ sage contains strong or weak arguments. When the elaboration likelihood is low, people know that they do not want and/or are not able to evaluate the merits of the arguments presented (or they do not even consider exerting effort to process the message). Thus, if any evaluation is formed at all, it is likely to be the result of relatively simple associations or inferences. When the elaboration likelihood is high, people know that they want to evaluate the merits of the arguments present­ ed and that they are able to do so. In these high-elaboration situations, persuasion­ relevant variables (such as source attractiveness) have relatively little impact by serving as simple cues. Instead, the variable can serve as a persuasive argument if it provides information central to the merits of the attitude object (e.g., a pic­ ture of an attractive source can provide persuasive visual testimony as to the ef­ fectiveness of a beauty product). Alternatively, if the variable is not of central relevance, it might bias the ongoing information processing activity (e.g., peo­ ple might generate mostly favorable thoughts about the message if the source is attractive). For example, consider the mere number of people who endorse a po­ sition. Under low-elaboration conditions, this variable would likely serve as a simple peripheral cue via the heuristic, “the more people who endorse the posi­ tion, the better.” However, under high-elaboration conditions, people might be motivated to think about why so many or so few people have endorsed the posi­ tion and generate arguments consistent with the direction of the cue (see Burnstein & Vinokur, 1975; Burnstein, Vinokur, & Trope, 1973). Thus, a consensus cue could lead to attitude change for very different reasons under high- and lowelaboration conditions. Finally, when the elaboration likelihood is moderate (e.g., uncertain personal relevance, moderate knowledge, moderate complexity, etc.), people might be un­ sure as to whether the message warrants or needs scrutiny, and whether they are capable of providing this analysis. In these situations, they might examine the persuasion context for indications of whether they should attempt to think about the message. For example, most people might be more likely to process the

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

99

message if it comes from an expert rather than a nonexpert source, because it often makes little sense to waste time thinking about a message from someone who does not know very much (Heesacker, Petty, & Cacioppo, 1983; Moore, Hausknecht, & Thamodaran, 1986). On the other hand, people who are high in dogmatism (Rokeach, 1960) simply might trust what an authority figure has to say without much scrutiny. In fact, research has indicated that high dogmatic in­ dividuals appear to accept a message from an expert source without much think­ ing about it, whereas low dogmatics are more dependent on the quality of the arguments provided (DeBono & Klein, 1993). The impact of expertise on the amount of elaboration should occur primarily in situations where the overall likeli­ hood of processing the message is neither very high nor low due to other factors. An important assumption of the ELM analysis is that the information process­ ing activity undertaken allows the person to satisfy his or her goals. In this regard, consider a voter who becomes interested in the political campaign quite late and is reading the pamphlets of the two candidates while in line waiting to vote. What if the person is unable to discriminate the candidates based on their issue posi­ tions? Earlier in the campaign, the voter might have been satisfied in viewing the candidates equally and hoping that some further substantive information would aid in the choice. However, at this point, the voter might use any positive cues available as a basis for an evaluation to break the deadlock (see Miniard, Sirdeshmukh, & Innis, 1992). Several studies have examined the impact of variables at two levels of elabo­ ration likelihood (high and low), and have provided evidence for a trade-off be­ tween argument processing and the impact of peripheral cue processes (Petty & Cacioppo, 1986a).14 Only one published study examined the effects of a vari­ able across three distinct levels of elaboration, however. This study (Moore et al., 1986; Experiment 3) provided support for the predictions outlined earlier (see Petty, Kasmer, Haugtvedt, & Cacioppo, 1987). Specifically, Moore et al. found that when an advertisement was presented at a very rapid pace, so that it was quite difficult to process, people were influenced greatly by the expertise of the product endorser, but the quality of the arguments for the product had lit­ tle effect. When the message was presented at a normal pace and was quite easy to process, the quality of arguments in the ad had more impact, but the expertise of the endorser has less influence than in the fast presentation condition. Finally, when the message was presented at a moderately fast pace and processing was possible but challenging, the expertise of the endorser determined how much processing occurred—the expert source induced more thinking than the non­ expert. 14There are a number of reasons why peripheral cues typically would have a smaller impact when the elaboration likelihood is high rather than low. For example, people might pay less attention to peripheral factors when they are motivated to scrutinize arguments, the cues may be evaluated less extremely, cues could become less salient at the time of judgment, or weighted less, and so on (e.g., McKenzie & Spreng, 1992; see Petty, in press, for discussion).

100

PETTY, PRIESTER, WEGENER

A number of persuasion variables in addition to source expertise are amena­ ble to a multiple roles analysis. For example, the ingroup or outgroup status of the message source serves as a simple acceptance/rejection cue when the mes­ sage is irrelevant to the in-group membership (i.e., low motivation to process), but influences the extent of message processing when the message has some relevance to the in-group status (Mackie, Worth, & Asuncion, 1990). To further explicate the multiple roles idea, we discuss two variables that have sparked con­ siderable research attention. Multiple Roles for Affect in Persuasion. A growing number of studies has examined the impact of a person’s mood during exposure to a communication on information processing and persuasion (see Breckler, in press; Mackie, Asun­ cion, & Rosselli, 1992; Petty et al., 1991; Schwarz, Bless, & Bohner, 1991; for reviews). According to the ELM, when the likelihood of issue-relevant thinking is low, a person’s mood will serve largely as a peripheral cue, providing mean­ ing to the attitude object by a simple association process. Consistent with this prediction, early investigations of mood and persuasion often were guided by clas­ sical conditioning notions of a direct association between the attitude object and the person’s affective state (e.g., Griffitt, 1970; Zanna, Kiesler, & Pilkonis, 1970). More recently, affective states have been postulated to influence attitudes by a simple inference process in which misattribution of the cause of the mood state to the persuasive message or to the attitude object occurs (e.g., I must feel good, because I like or agree with the message advocacy; see Cacioppo & Petty, 1982; Petty & Cacioppo, 1983; Schumann & Thorson, 1990; Schwarz, 1990). These direct effects of mood on attitude seem to be more likely when elaboration likeli­ hood is low than high (Gom, 1982; Miniard, Bhatla, & Sirdeshmukh, 1992; Srull, 1983). As the likelihood of elaboration increases, mood takes on different roles. Spe­ cifically, when the elaboration likelihood is more moderate, mood (like source attractiveness, Puckett, Petty, Cacioppo, & Fisher, 1983; and credibility, Moore et al., 1986) has been shown to have an impact on the extent of argument elabo­ ration. In particular, people who have been placed in a positive mood have shown less inclination to process message arguments than people in a neutral or nega­ tive mood (Batra & Stayman, 1990; Bless, Bohner, Schwarz, & Strack, 1990; Kuykendall & Keating, 1990; Mackie & Worth, 1989; Worth & Mackie, 1987). Competing theoretical positions have been put forward to explain why posi­ tive mood disrupts processing. The “cognitive capacity” view states that positive moods activate many positive thoughts in memory. Because these thoughts oc­ cupy a person’s attentional capacity, the message recipient becomes less able to process incoming information (Mackie & Worth, 1991). Support for this formu­ lation comes from the finding that processing deficits for people in a positive mood can be eliminated by removing time pressures for processing (e.g., Mackie & Worth, 1989). However, in some studies where time to process the persuasive

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

101

messages was unlimited but not emphasized, processing deficits still have been found for happy people (e.g., Batra & Stayman, 1990; Kuykendall & Keating, 1990). Thus, it could be that both a motivational as well as an ability process is operating to reduce elaboration by people in a positive mood. Or, as noted by Schwarz et al. (1991), the time pressure manipulation in the Mackie and Worth (1989) studies might be interpreted as varying subjects’ motivation to process the message. One motivational account of mood effects on message processing is the “feelings-as-information” view (Schwarz, 1990). According to this framework, negative affective states signal to the message recipient that something is wrong in the environment and that some action is necessary, whereas positive mood in­ dicates that no scrutiny of the environment is required. Although the Bless et al. (1990) studies support such a contention, other studies have been less con­ genial to the feelings-as-information position. For instance, some persuasion in­ vestigators have argued that positive mood enhances thinking, based on the fact that happy mood has increased memory for a message over sad mood (Goldberg & Gorn, 1987), and happy individuals have generated a greater proportion of content-related cognitive responses than sad individuals (Mathur & Chattophadhyay, 1991). The cognitive capacity explanation and the feelings-as-information analysis both predict that positive mood disrupts processing. Therefore, to the extent that increases in processing can be found in positive moods relative to neutral or sad moods (Isen, Daubman, & Nowicki, 1987; Murray, Sujan, Hirt, & Sujan, 1990), alternative conceptual frameworks would need to be developed. One possible alternative relies on the notion of mood management (e.g., Sinclair & Mark, 1992). That is, positive-mood enhancement of processing might occur primarily when processing allows the message recipient to maintain his or her positive mood, whereas processing deficits might occur when processing would make positive moods less likely to persevere (for similar ideas regarding mood and helping, see Isen & Simmonds, 1978). When the elaboration likelihood is high and people are processing the mes­ sage arguments already, the ELM holds that affective states can influence atti­ tudes by influencing the nature of the thoughts that come to mind. Positive mood can facilitate the retrieval of positive and/or inhibit the retrieval of negative material from memory (e.g., see Blaney, 1986, Bower, 1981; Clark & Isen, 1982). Thus, a person’s mood during message processing should be related to the favorability of the cognitive responses generated. Support for this possibility has been found in a number of studies (e.g., Breckler & Wiggins, 1991; Goldberg & Gorn, 1987; Mathur & Chattopadhyay, 1991). Additional support has come from research in which the elaboration likelihood was varied explicitly. Specifically, in two studies (Petty, Schumann, Richman, & Strathman, 1993) positive mood had an impact on the thoughts and attitudes of subjects high in elaboration likelihood (i.e., for people high in need for cognition or under high message relevance con­ ditions), but mood influenced only attitudes when the elaboration likelihood was

102

PETTY, PRIESTER, WEGENER

low. Furthermore, under high-elaboration conditions, positive mood had an in­ direct effect on attitudes, being mediated by the positivity of the cognitive responses generated. However, under low-elaboration conditions, a direct effect of mood on attitudes was observed that was not mediated by thought positivity. As noted earlier, the biasing effect of mood on cognitive responses observed under highelaboration conditions should be most pronounced when the message arguments are relatively ambiguous and open to multiple interpretations. We do not mean to imply that positive moods always should bias processing toward being more favorable toward the position advocated or that negative moods invariably should render the message conclusion less acceptable. For instance, negative moods have been shown to make negative events seem more likely than positive events (Johnson & Tversky, 1983) and positive events and behaviors are seen as more likely in positive as opposed to negative moods (e.g., Erber, 1991; Mayer, Gaschke, Bravermen, & Evans, 1992). Thus, to the extent that message arguments include statements that a plan should be followed to avoid negative consequences, negative moods might predispose people to view these arguments as more compelling (see Petty & Wegener, 1991 for a discussion). For example, if a message recipient experiences fear during a message on cigarette smoking, this might influence the person’s perception of the severity of the threat or the subjective likelihood of the negative consequences postulated (Rogers, 1983; Schwarz, Servay, & Kumpf, 1985).15 On the other hand, if the message recipient is in a positive mood, perceptions of the likelihood of the favorable consequences in the message would be inflated (Wegener, Petty, & Klein, in press). Multiple Roles for Majority and Minority Sources. Like many persuasion vari­ ables, initial inquiries into the effects of majority/minority source status tended to ask main effect questions, such as whether majorities or minorities had greater influence (Asch, 1956; Moscovici, Lage, & Naffrechoux, 1969; Tanford & Pen­ rod, 1984) or whether majorities and minorities induced persuasion by the same or different processes (Moscovici, 1980). However, it has been acknowledged more recently that the majority /minority status of a source can have various ef­ fects depending on the configuration of motivational and cognitive factors present (Kruglanski & Mackie, 1990; Mugny & Perez, 1991). This complexity can be seen in studies that have investigated the amount of message processing fostered by majority as opposed to minority sources (Maass & Clark, 1983; Mackie, 1987) and by discussion of how source status can bias message processing (e.g., Kruglan­ ski & Mackie, 1990; Mackie, 1987). The multiple roles postulate of the ELM suggests that the different roles for majority and minority sources are more like­ ly to occur under some conditions than others. l5Like other affects, fear could serve multiple roles. For example, in addition to biasing elabo­ ration, fear appears to reduce the overall likelihood of message elaboration, especially when the fear related to the message topic is chronic (Jepson & Chaiken, 1990), or when there is a strong reassur­ ance that a proposed action will be effective (Gleicher & Petty, 1992).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

103

Specifically, when the elaboration likelihood is low, majority/minority source status is most likely to serve as a simple persuasion cue. That is, source status, by influencing perceptions of source credibility, competence, or trustworthiness, can provide message recipients with a simple decision rule as to whether they should agree with the message. However, source perceptions do not necessarily covary directly with majority/minority status (Kruglanski & Mackie, 1990). For instance, consistency in a source’s behavior can increase attributions of compe­ tence and thereby increase influence (Bray, Johnson, & Chilstrom, 1982; Maass, Clark, & Haberkorn, 1982; Moscovici & Neve, 1973); or can increase attribu­ tions of rigidity, and thereby decrease influence (e.g., Levine, Saxe, & Harris, 1976; see Maass & Clark, 1984). When the elaboration likelihood is moderate, majority/minority status can determine the amount of message processing that occurs. However, researchers have reached conflicting conclusions regarding whether majority sources or minority sources foster greater message elaboration. For instance, Mackie (1987) found that subjects tended to recall more majority than minority arguments, and they also generated more favorable cognitive responses to the majority message. Based on these findings, she concluded that majority sources induce greater message processing than minorities do. Other researchers have con­ cluded that the opposite is the case (e.g., Maass & Clark, 1983). It is quite possible that either majorities or minorities can induce greater message scrutiny, depending on other factors in the persuasion situation. For example, receiving a counterattitudinal message from a majority source implies that the message recipient is in the minority. This might be surprising and even threatening to the message recipient, making scrutiny of what the majority has to say more likely than might normally be the case. However, if a recipient encounters a proattitudinal message, the same reasoning predicts greater processing of that message when the source is a minority rather than a majority, because a proattitudinal message from a minority can be surprising and/or threatening (Baker & Petty, 1993). Finally, when the motivation and ability to process an incoming message are high, majority/minority status should impact persuasion primarily by influenc­ ing the nature of the processing that occurs. For example, Mackie (1987) noted that, in her research, biased processing might have produced the relatively great­ er proportion of positive cognitive responses that subjects generated to messages from the majority source. Trost, Maass, and Kenrick (1992) found that, although a minority message was persuasive when its personal relevance was low, recipients for whom the message was highly relevant derogated the minority message more than the majority communication. Thus, biased information processing of majority and minority sources might be exacerbated under conditions that foster elabora­ tion. The majority/minority status of a source also can act as an argument in highelaboration conditions. For instance, if a message recipient encounters a mes­ sage about which position is likely to be supported in a local election, the number

104

PETTY, PRIESTER, WEGENER

of people who currently support the position is a piece of information that is directly relevant to the central merits of the position being presented.16 Peripheral Mechanisms in Attitude Change

In the central route to persuasion, the presumption is that attitude change results from actively considering the merits of the position being advocated. That is, people receive and elaborate arguments or generate reasons to explain or justify some outcome or behavior. Sometimes this effortful thinking proceeds in a rela­ tively objective manner, but at other times it clearly is biased. In either case, however, the person is engaged in an active processing of the issue-relevant in­ formation presented. On the other hand, attitudes also can be changed as a result of various peripheral cues in the persuasion setting if the motivation or ability to elaborate is low. The theories that we discuss next postulate a number of specific peripheral processes. Although many of these theories propose that the processes involved are quite general, subsequent research has revealed that they tend to operate most effec­ tively when the likelihood of issue-relevant elaboration is low or the attitude to be modified is relatively weak. We begin with theories that emphasize inference and heuristic processes, and conclude with theories that emphasize the associa­ tion of affect with attitude objects. Inference-Based Approaches

Attribution Theory. The 1970s brought an explosion of interest within psy­ chology in examining how people came to understand the causes of their own and others’ behavior. The gist of this attributional approach was that people came to infer underlying characteristics about themselves and others from the behaviors that they observed and the situational constraints imposed on these behaviors (Jones & Davis, 1965; Kelley, 1967). In a provocative paper, Bern (1965) suggested that people sometimes have no special knowledge of their own internal states and simply infer their attitudes in a manner similar to that by which they infer the attitudes of others. In his selfperception theory, Bern reasoned that just as people assume that the behavior of others and the context in which it occurs provides information about the presumed 16Space limitations preclude discussion of other variables that have been shown to serve in mul­ tiple roles. For example, a person’s own attitude provides a ready cue for evaluation when the likeli­ hood of thinking is low (Jamieson & Zanna, 1989; Sanbonmatsu & Fazio, 1990), but when the likelihood of thinking is high, attitudes bias the ongoing information-processing activity (Houston & Fazio, 1989; Liberman & Chaiken, 1992; see Petty et al., 1992). Both the cue and biasing effects are more likely the more accessible the attitude is (Fazio, in press). Finally, people also might be more interested in objects about which they have attitudes (Roskos-Ewoldson & Fazio, 1992b), and thus attitudes also might determine the extent of information processing activity.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

105

attitudes of these people, so too would a person’s own behavior provide informa­ tion about the person’s own attitude. During much of the 1970s, self-perception theory was thought to provide an alternative account of dissonance effects (Bern, 1972). However, subsequent research indicated that both dissonance and self­ perception processes can operate, but in different domains. In particular, the un­ derlying tension mechanism of dissonance theory operates when a person engages in attitude-discrepant action that is disagreeable (e.g., advocating a discrepant position in one’s latitude of rejection, Fazio, Zanna, & Cooper, 1977; perform­ ing self-deprecating behavior, Jones, Rhodewalt, Berglas, & Skelton, 1981), whereas self-perception processes are more likely when a person engages in attitude-discrepant, but more agreeable, behavior (e.g., advocating a discrepant position in one’s latitude of acceptance, Fazio et al., 1977; performing a self­ enhancing behavior, Jones et al., 1981). Self-perception theory also accounted for some unique attitudinal phenome­ na. For example, the overjustification effect occurs when a person is provided with more than sufficient reward for engaging in an action that is already highly regarded (e.g., Crano, Gorenflo, & Shakelford, 1988; Lepper, Greene, & Nisbett, 1973). To the extent that the person comes to attribute the action to the ex­ ternal reward, rather than to the intrinsic enjoyment of the behavior, attitudes toward the behavior will become less favorable (Deci, 1975). Thus, if people are provided with extrinsic rewards for advocating a position that they already like, they may come to devalue the position when the external rewards stop, be­ cause they have come to view their attitude expression as caused by the rewards, rather than by the true merits of the position (e.g., Scott & Yalch, 1978). According to the ELM, people should be more likely to rely on this relatively simple inference process when well-defined attitudes are not very accessible or the elaboration likelihood is low (Tybout & Scott, 1983). For example, Wood (1982) examined the power of self-perception processes for people who had rela­ tively high versus low knowledge and experience with the issue of environmen­ tal preservation. Subjects committed themselves to deliver a speech that was consistent with their attitudes after learning that they would receive either $5.00 or nothing for the task. Following this, they expressed their opinions on the is­ sue. The major result was that for subjects with low knowledge and experience, the $5.00 incentive undermined their positive attitudes (e.g., “I must have made the pro-environmental statements for the money”), but for high knowledge sub­ jects, the incentive had no effect (see also Chaiken & Baldwin, 1981). Finally, the attribution approach also has been useful in understanding how people make inferences about relatively simple cues. For example, Eagly, Chai­ ken, and Wood (1981) argued that people often approach a persuasion situation with some expectation regarding the communicator’s position. This expectation is formed from premessage information regarding the communicator’s traits and situational pressures. According to this model, if the premessage expectation about the position to be taken is confirmed by the communicator’s presentation, little

106

PETTY, PRIESTER, WEGENER

persuasion occurs, because the recipient attributes the message to the traits and pressures that generated the expectation. In these cases, the person needs to process the message to determine its validity. However, when the premessage expecta­ tion is discontinued, the communicator is seen as relatively accurate and unbi­ ased, and persuasion can be increased without the need to process message arguments (Eagly, Wood, & Chaiken, 1978). On the other hand, suppose disconformation of expectations leads to an attribution of untrustworthiness, rather than honesty. In these situations, the processing consequences of disconfirma­ tion are reversed (Priester & Petty, 1991). In addition, differential processing consequences of expectancy confirmation and disconfirmation are especially likely for individuals who are inclined to conserve their cognitive resources (i.e., those low in need for cognition; Priester & Petty, 1993).17 The Heuristic/Systematic Model. Like the attributional framework developed by Eagly and her colleagues, Chaiken’s heuristic/systematic model of persuasion (HSM) represents an explicit attempt to explain why certain peripheral cues, such as source expertise or message length, have the impact that they do (Chaiken, 1987; Chaiken et al., 1989). However, the HSM focuses on heuristics retrieved from memory, rather than decision rules that are generated on-line. That is, Chai­ ken proposed that, in contrast to systematic (or central route) processes, many source, message, and other cues are evaluated by means of simple schemas or cognitive heuristics that people have learned on the basis of past experience and observation. To the extent that various persuasion rules of thumb are available in memory and accessed, they can be invoked to evaluate persuasive communi­ cations. According to the HSM, the likelihood of systematic processing increases whenever confidence in one’s attitude drops below the desired level of confidence (the sufficiency threshold). Whenever actual and desired confidence are equal, heuristic processing is more likely. For example, because of either prior per­ sonal experience or explicit training, people could base their acceptance of a mes­ sage on the mere number of arguments contained in it by invoking the heuristic “the more arguments in favor of something, the more valid it is” (a length implies strength heuristic; Alba & Marmorstein, 1987; Petty & Cacioppo, 1984a; Wood et al., 1985). The HSM holds that, as the elaboration likelihood is increased, there is an increased likelihood of heuristic and systematic information process­ ing. In particular, as long as the two processing modes do not yield conflicting reactions (such as when an expert source presents weak arguments), heuristic

17If attributional reasoning were employed in the service of evaluating the specific attributes of a position, rather than providing a simple inference about the communicator or other cue, the result­ ing attitude would have more in common with central than peripheral route attitudes.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

107

processing adds to the impact of whatever systematic processing has occurred (Chaiken et al., 1989).18 Although some research has varied the accessibility (Roskos-Ewoldsen & Fa­ zio, 1992a) or vividness (Pallak, 1983) of the peripheral cues in a message, rela­ tively little research has attempted to explicitly test the HSM by varying the accessibility of the heuristics. Chaiken (1987) reported the most pertinent inves­ tigations. In four studies, she and her colleagues attempted to make certain deci­ sion rules more accessible, and then track their influence on attitudes following message exposure. Although the individual studies produced weak results, the general pattern supported the utility of the heuristic idea. For example, in one study, subjects in the experimental condition memorized eight phrases relevant to the length implies strength heuristic (e.g., “the more the merrier”), whereas control subjects memorized eight irrelevant phrases. Subsequently, subjects received a message from a speaker who claimed to have either 10 or 2 reasons in support of mandatory comprehensive exams for seniors. Subjects in the study also were divided into those with high and low need for cognition (Cacioppo & Petty, 1982). The only group to be influenced significantly by the claim of 10 versus 2 reasons was composed of low need for cognition subjects who had been primed with the relevant phrases. Cialdini (1987) analyzed several other heuristics that are effective in influencing behavior. For example, although people might reason that the more people endors­ ing an object, the better it is, they also can reason that the fewer the number of ob­ jects available, the better it is (i.e., the scarcity heuristic; see Brock, 1968). Of course, many of these variables should influence attitudes by invoking heuristics mostly when the elaboration likelihood is low. As noted previously, as the elabora­ tion likelihood is increased, these variables take on other roles (e.g., perceived scarcity can increase the extent of message processing; Bozzolo & Brock, 1992). Approaches Emphasizing Affect

The attribution and heuristic models focus on simple cognitive inferences that can modify attitudes. Next, we discuss theoretical approaches emphasizing the role of relatively simple affective processes in attitude change. 18Although the ELM proposes a trade-off between the impact of central and peripheral process­ es along the elaboration likelihood continuum, this does not mean that central and peripheral processes cannot co-occur (cf. Petty et al., 1987; Stiff, 1986). In fact, at most points along the continuum, some joint impact would be expected. Furthermore, the ELM does not hold that peripheral processes are less likely to occur as the elaboration likelihood is increased—only that these peripheral process­ es are less likely to have an impact on the person’s attitude. Finally, the ELM does not imply that variables serving as peripheral cues cannot have an impact on attitudes as the elaboration likelihood is increased. Rather, the multiple-roles postulate argues that the same variable can have an impact under high- and low-elaboration conditions, although the mechanism would be different (see Petty, in press; Petty, Wegener, Fabrigar, Priester, & Cacioppo, in press, for additional discussion).

108

PETTY, PRIESTER, WEGENER

Classical Conditioning. One of the most direct means of associating affect with objects, issues, or people is through classical conditioning. In brief, condi­ tioning occurs when an initially neutral stimulus (the conditioned stimulus; CS) is associated with another stimulus (the unconditioned stimulus; UCS) that is con­ nected directly or through prior conditioning to some response (the unconditioned response; UCR). By pairing the UCS with the CS, the CS becomes able to elicit a conditioned response (CR) that is similar to the UCR (see Gardner, 1985; McSweeney & Bierley, 1984; Petty, Cacioppo, & Kasmer, 1988, for relevant reviews). Considerable psychological research has shown that attitudes can be modified by pairing initially neutral objects with stimuli about which people already feel positively or negatively. For example, people’s evaluations of words (e.g., Staats & Staats, 1958), other people (e.g., Griffit, 1970), political slogans (e.g., Razran, 1940), products (e.g., Gresham & Shimp, 1985), and persuasive communi­ cations (e.g., Rogers, 1983) have been modified by pairing them with such affect-producing stimuli as unpleasant odors and temperatures, the onset and off­ set of electric shock, harsh sounds, pleasant pictures, and elating and depressing films (e.g., Gouaux, 1971; Staats, Staats, & Crawford, 1962; Stuart, Shimp, & Engle, 1987; Zanna et al., 1970). Various motoric movements that are associat­ ed with positive and negative consequences, such as putting one’s face in a smil­ ing versus a frowning pose (Strack, Martin, & Stepper, 1988) and making arm flexing movements (associated with moving objects toward oneself), rather than extension movements (Cacioppo, Priester, & Bemtson, 1993), also can influence evaluative responses. As noted in our discussion of multiple roles for affect, people should be espe­ cially susceptible to simple conditioning (and other peripheral) processes when the likelihood of object-relevant thinking is rather low. For example, in one per­ tinent experiment, subjects were exposed to initially neutral words and nonwords while receiving electric shock either immediately after the words, immediately after the nonwords, or randomly (Cacioppo, Marshall-Goodell, Tassinary, & Petty, 1992). Conditioning was more effective for the items (i.e., nonwords) for which subjects had no preexisting knowledge or associations (see also, Shimp, Stuart, & Engle, 1991). In another relevant study (Mackie & Worth, 1990), a persuasive message was presented to subjects over headphones on one channel while either positive (e.g., happiness) or negative (e.g., pain) distraction words were presented on another. Subjects were instructed that attending to the mes­ sage was the primary task. The message first provided information about a con­ sumer product (e.g., a soda) and described favorable attributes that were either central to the merits of the product (e.g., good taste, thirst quenching) or rather peripheral (e.g., reputable brand name). Following this, subjects received addi­ tional information about the product. The major result was that the valence of the distraction words presented in the unattended ear influenced the individuals whose product attitudes were based on the peripheral information to a greater

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

109

extent than the individuals whose attitudes were based on more important infor­ mation. Thus, the affectively laden distraction words had a greater impact on the weaker attitudes.19 Affective Priming. In a procedure that bears some similarity to classical con­ ditioning, subjects are presented with affect-inducing material just prior to receipt of the target stimulus so that positive or negative affect is primed and might there­ fore influence reactions to subsequent stimuli. This “backward conditioning” or “affective priming” procedure has proven successful in modifying attitudes. In one study, for example, subjects were shown a series of nine photos of a target person going about normal daily activities (e.g., getting into a car; Krosnick, Betz, Jussim, & Lynn, 1992). Just prior to each picture of the target person, sub­ jects were exposed to a subliminal photo that was pretested to elicit positive (e.g., a group of smiling friends) or negative (e.g., a bucket of snakes) affect. Subjects exposed to the positive subliminal slides rated the target person more favorably than subjects exposed to the negative slides. In a series of studies, Murphy and Zajonc (1993) found that the effectiveness of the affective priming procedure may be dependent on presenting the primes outside of conscious awareness. For example, when positive and negative affec­ tive primes (smiling and frowning faces) were presented just prior to a target stimulus (a Chinese ideograph), attitudes toward the target were influenced only when the primes were presented subliminally (a 4 ms backward masking presen­ tation) and not when they were presented visibly (a 1 s presentation). When the primes were available for conscious analysis, features other than their primitive affective qualities (e.g., how physically attractive the pictured individual was) were invoked and this attenuated the affective priming effect. Mere Exposure. Another procedure for modifying attitudes through simple affective means was identified by Zajonc (1968a). In this research, Zajonc and his colleagues showed consistently that when objects simply are presented to the individual on repeated occasions, this mere exposure is capable of making the individual’s attitude toward these objects more positive (Zajonc & Markus, 1982). An early explanation of the mere exposure effect was provided by Titchener (1910), who proposed that familiar objects led people to experience a “glow of warmth, a sense of ownership, a feeling of intimacy” (p. 411). The most recent work on this phenomenon indicates that simple repetition of objects can lead to more positive evaluations even when people do not recognize that the objects are familiar. For example, in one study, Kunst-Wilson and Zajonc 19Mackie and Worth (1990) interpreted their results in terms of Fiske and Pavelchak’s (1986) category/piecemeal model. That is, they argued that the evaluations of products based on important or prototypical attributes (category based evaluation) required less cognitive capacity than evalua­ tions based on the unimportant or atypical attributes (piecemeal evaluation), and therefore the former attitudes were less vulnerable to influence from the distracting material.

110

PETTY, PRIESTER, WEGENER

(1980) visually presented polygon images to subjects a number of times under viewing conditions that resulted in chance reports of recognition. During a later session, subjects were shown pairs of polygons under ideal viewing conditions. In each pair, one shape had been seen in the earlier session, but the other was new. Subjects were asked which shape they liked better and which one they had seen before. Although subjects were unable to recognize beyond chance which of the polygons was new and which was old, they showed a significant prefer­ ence for the “old” over the new shapes. Mere exposure effects have been shown in a number of studies using a variety of stimuli in addition to polygons such as tones, nonsense syllables, Chinese ideo­ grams, photographs of unfamiliar faces, and foreign words (e.g., Matlin, 1970; Wilson, 1979; Zajonc, 1968a; see Bornstein, 1989, for a review). All of these stimuli tend to be low in meaning, and thus are relatively unlikely to elicit spon­ taneous elaboration. In fact, the simple affective process induced by mere ex­ posure appears to be especially successful in influencing attitudes when conscious processing of the repeated stimuli is minimal (Obermiller, 1985) or impossible such as when the stimuli are presented subliminally (Bornstein & D’Agostino, 1992). When more meaningful stimuli have been repeated, such as words or sen­ tences, mere exposure effects have been less common. Instead, when meaning­ ful processing occurs with repetition, the increased exposures enhanced the dominant cognitive response to the stimulus. Thus, attitudes toward negative words (e.g., hate) and weak message arguments become more unfavorable, but atti­ tudes toward positive words (e.g., love) and strong arguments become more favorable, at least until the point of tedium (e.g., Cacioppo & Petty, 1985, 1989; Grush, 1976; Sawyer, 1981). In sum, studies of affect association and mere exposure show that primitive affective processes are most likely to influence attitudes for low knowledge, low relevance, and/or initially meaningless attitude objects or issues. However, this does not mean that affect will influence attitudes only when the elaboration likeli­ hood is low. As noted in our discussion of multiple roles, affect can modify atti­ tudes when the elaboration likelihood is higher. However, the underlying processes of change are different. Information Integration Stage

Once an individual has received and accepted some portion of the message argu­ ments, generated idiosyncratic cognitive responses to them, made attributions, retrieved heuristics elicited by various cues in the persuasion setting, or ex­ perienced a number affective states, the information units and experiences must be combined in some manner to form an overall attitudinal reaction. Two com­ binatory rules have generated the bulk of research attention. Fishbein and Ajzen’s (1981) expectancy-value formulation, described earlier, provides one such

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

111

model. Although this model has been applied mostly to evaluations of issue­ relevant arguments, it is potentially applicable to integrating more peripheral fea­ tures of the communication as well. According to this model, following exposure to a persuasive communication, the object-relevant attributes (whether provided in the message or recipient generated) that are salient are combined in the man­ ner specified by Equation 4 (Fishbein & Ajzen, 1981). An alternative integration formula is provided by Anderson’s (1971) informa­ tion integration theory. In contrast to the additive rule specified by Fishbein and Ajzen, Anderson posited that the pieces of information in a communication (or one’s cognitive responses; Anderson, 1981) typically are combined by an aver­ aging process. Specifically, the person’s evaluation of the salient information (scale value = s,) is weighed by the importance of the information for the judgment (weight = wt) and is averaged with the person’s weighed (w0) initial attitude (sa) to form a new attitude (A0). A0 = (w0s0 + I W&) / (wQ + I wt)

(5)

Most tests of this formula have relied on topics and issues for which the recipients have relatively little prior information, such as the qualities of deceased presidents (e.g., Theodore Roosevelt; Sawyers & Anderson, 1971) and evalua­ tions of hypothetical cars (Birnbaum & Stegner, 1979). The model has performed well in these settings. Perhaps the biggest strength of this model is simultaneous­ ly its greatest weakness, namely, the very flexible use of the weighting parameter (Eagly & Chaiken, 1984; Petty & Cacioppo, 1981). On the positive side, this parameter can be used to model the impact of many variables, such as source credibility (Birnbaum, Wong, & Wong, 1976) and multiple sources (Himmelfarb, 1972). On the other hand, the formula says little a priori about when (or why) these variables affect the impact of the information. One clear implication of the averaging model of information integration is that it implies that it should be more difficult to change attitudes the greater their in­ formational base. There are two reasons for this. First, the more relevant beliefs that are retrieved at the time of judgment to integrate with new beliefs, the less the new beliefs will impact on the overall attitude, because of averaging. Second, even if a person does not retrieve the specific beliefs associated with the attitude at the time of judgment, prior attitudes based on extensive beliefs are expected to be given greater weight than attitudes based on few beliefs (Anderson, 1981; Davidson, in press). This, too, would render any new information less impactful on attitudes. Thus, according to Equation 5, one can use processes that occur during the information integration stage of information processing to account for some instances of attitude strength, such as resistance to change, rather than relying on processes that occur during yielding (e.g., enhanced counterarguing due to prior beliefs).

112

PETTY, PRIESTER, WEGENER

CONSEQUENCES OF DIFFERENT ATTITUDE CHANGE PROCESSES

Our review of the major theories of persuasion focused on the processes respon­ sible for yielding to a persuasive communication. In addition, we used the ELM to suggest some of the general conditions under which various categories of processes would be most likely to operate. For example, consider the decadelong controversy regarding the viability of dissonance versus self-perception the­ ories. Both conceptualizations predicted the same pattern of attitude results in many situations, although for different reasons (Greenwald, 1975). As noted, self­ perception theory focused on a relatively simple cognitive inference as the basis of change, whereas dissonance theory postulated a more cognitively active process of rationalization. Thus, the dissonance process should be more likely when con­ ditions fostered a high likelihood of elaboration (i.e., conditions of high personal relevance, consequences, responsibility, prior knowledge, etc.), rather than low. Research is generally supportive of this view (Cooper & Fazio, 1984; Fazio et al., 1977). On the other hand, self-perception processes should be more likely to operate when the elaboration likelihood is rather low (i.e., conditions of low personal relevance, consequences, responsibility, knowledge, etc.). As noted previously, this also has received empirical support (Chaiken & Baldwin, 1981; Taylor, 1975; Wood, 1982). In this section, we turn briefly to some of the important characteristics and consequences of attitudes that are changed by different processes. In particular, we focus on the temporal persistence of attitude changes, the resistance of atti­ tude changes to counterpersuasion, and the ability of newly formed or changed attitudes to predict behavior (and behavioral intentions). As noted earlier, each of these qualities is indicative of attitude strength. Over the past few decades, various studies have addressed these issues. A general conclusion is that attitude changes that are accompanied by high levels of issue-relevant cognitive activity about the dimensions central to the attitude object are stronger than changes that are accompanied by little issue-relevant thought, or considerable thought, but along dimensions that are not central to the merits of the attitude object (Petty, Haugtvedt, & Smith, in press). High levels of issue-relevant cognitive activity are like­ ly to require frequent accessing of the attitude and the corresponding knowledge structure. Therefore, this activity should tend to increase the number of linkages and strengthen the associations among the structural elements, making the atti­ tude schema more internally consistent, accessible, and enduring (Crocker, Fiske, & Taylor, 1984; Fazio, et al., 1986; McGuire, 1981). In comparison, attitude change that results from a simple on-line inference or heuristic process typically involves accessing the attitude structure only once to incorporate the affect or inference associated with a salient persuasion cue (Petty & Cacioppo, 1986a). In general, then, these attitudes should be weaker.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

113

Persistence of Attitude Change

Persistence of persuasion refers to the extent to which an attitude change endures over time. In a comprehensive review of the experimental work on attitude per­ sistence, Cook and Flay (1978) concluded quite pessimistically that most of the laboratory studies on attitude change tended to find very little persistence of per­ suasion. In the years since this influential paper, it has become more clear when attitude changes will persist and when they will not. In particular, current research has indicated that when attitude changes are based on extensive issue-relevant thinking, they tend to endure (e.g., Mackie, 1987). That is, conditions that foster people’s motivation and ability to engage in issue-relevant cognitive activity at the time of message exposure are associat­ ed with increased persistence of persuasion. For example, research has shown that self-generation of arguments (Elms, 1966; Watts, 1967) and autobiographi­ cal instances relevant to an issue (Lydon, Zanna, & Ross, 1988), the use of in­ teresting or involving communication topics (Ronis, Baumgardner, Leippe, Cacioppo, & Greenwald, 1977), providing increased time to think about a mes­ sage (Mitnick & McGinnies, 1958), increasing message repetition (Johnson & Watkins, 1971), reducing distraction (Watts & Holt, 1979), and leading recipients to believe that they might have to explain or justify their attitudes to other people (Boninger, Brock, Cook, Gruder, & Romer, 1990; Chaiken, 1980) are all as­ sociated with increased persistence. Also, people who characteristically enjoy thinking (high need for cognition) show greater persistence of attitude change than people who do not (Haugtvedt & Petty, 1992; Verplanken, 1991). Simple cues can become associated with persistent attitudes if the cues remain salient over time. This can be accomplished by repeated pairings of the cue and attitude object so that the cue remains relatively accessible (Weber, 1972) or by reintroducing the cue at the time of attitude assessment (Kelman & Hovland, 1953). The Yale group explicitly acknowledged the potential impact of peripheral cues in their work on the sleeper effect. A sleeper effect is said to occur when a mes­ sage that is accompanied initially by a negative cue (e.g., a low credible source) increases in effectiveness over time (Cook, Gruder, Hennigan, & Flay, 1979; Hovland, Lumsdaine, & Sheffield, 1949). To account for this effect, Kelman and Hovland (1953) proposed that, in addition to message arguments, various cues could have an impact on attitude change. These cues were thought to add to (or subtract from) the effects of the persuasive message. In this model, the cues and message were viewed as independent and were postulated to have different decay functions. Given this formulation, a sleeper effect would be produced if a person were exposed to a message with a discount­ ing cue and the following conditions were met: (a) the message alone had a strong positive impact, (b) the discounting cue was sufficiently negative to suppress the positive impact of the message, and (c) the message conclusion became disso­ ciated from the discounting cue more quickly than it became dissociated from the

114

PETTY, PRIESTER, WEGENER

message arguments (Cook et al., 1979). Thus, at a later point in time, it is possi­ ble for the positive residue of the message to outlast the negative effect of the cue, leading to increased agreement with the message conclusion. The conditions that are postulated to produce a sleeper effect are unique and have several implications. In particular, the discounting cue formulation requires that the effects of a low credible source and strong message arguments operate jointly and additively. Yet, as reviewed, considerable research has shown that there is often a trade-off between the operation of source factors and the impact of message arguments, or an interaction between them. When the elaboration likelihood is high, arguments tend to be the primary determinant of persuasion; but when the elaboration likelihood is low, source and other cues tend to dominate. When the elaboration likelihood is moderate, the nature of the source could de­ termine the extent of argument processing. In the 20 years after the Kelman and Hovland (1953) study, so few sleeper effects appeared in the literature that some researchers declared that the effect did not exist (Gillig & Greenwald, 1974). One key to producing a sleeper effect is to construct a situation in which both a strong negative cue and strong argu­ ments have an initial impact. However, as noted, this should be difficult to produce because of the trade-off between the impact of arguments and cues, or an inter­ action between them (e.g., people might ignore a message from a low credible source, overturning one of the critical conditions for the effect). A clever solu­ tion to this dilemma is to have subjects process the message so that the strength of the arguments is realized, and following this present a discounting cue that causes subjects to doubt the validity of the message. This is the procedure em­ ployed in a number of successful sleeper effect studies (Cook et al., 1979; Kel­ man & Hovland, 1953; Mazursky & Schul, 1988). In a relevant series of experiments, Pratkanis, Greenwald, Leippe, and Baumgardner (1988) showed that presenting the discounting cue after the message might be critical for obtain­ ing a reliable sleeper effect.20 Resistance to Counterpersuasion

Resistance refers to the extent to which an attitude change is capable of surviving an attack from contrary information. Attitudes are more resistant the less they change in the direction of contrary information. Although attitude persistence and resistance tend to co-occur, their potential independence was shown conclusive­ ly in McGuire’s (1964) work on cultural truisms. Truisms such as “you should brush your teeth after every meal” tend to be highly persistent in a vacuum, but very susceptible to influence when challenged. As McGuire noted, people have 20In some research, it is difficult to determine if a true sleeper effect was obtained or whether the increase in persuasion over time is a result of decay of “boomerang” (e.g., Hannah & Stemthal, 1984).

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

115

very little practice in defending these beliefs, because they never have been at­ tacked. These beliefs likely were formed with little issue-relevant thinking at a time during childhood when extensive thinking was relatively unlikely. Instead, the truisms probably were presented repeatedly by powerful, likable, and expert sources. As noted, the continual pairing of a belief with positive cues can produce a relatively persistent attitude, but these attitudes might not prove resistant when attacked. The resistance of attitudes can be improved by bolstering them with relevant information (Lewan & Stotland, 1961). In his work on inoculation theory, McGuire (1964) demonstrated that two kinds of bolstering can be effective in inducing resistance. One form relies on providing individuals with a supportive defense of their attitudes or having them generate supportive information. For example, subjects whose initial attitudes were bolstered by recalling autobiographi­ cal instances relevant to the attitude showed greater resistance to an attacking message than subjects whose attitudes were followed by the generation of au­ tobiographical instances that were irrelevant to the attitude issue (Ross, McFarland, Conway, & Zanna, 1983). A second type of defense relies on a biological analo­ gy. That is, McGuire suggested that just as people can be made more resistant to a disease by giving them a mild form of the germ, people could be made more resistant to discrepant messages by inoculating their initial attitudes. The inocu­ lation treatment consists of exposing people to a few pieces of counterattitudinal information prior to the threatening communication and showing them how to refute this information. This presumably produces subsequent resistance, because the inoculation poses a threat that motivates and enables people to develop bol­ stering arguments for their somewhat weakened attitude (see also, McGuire & Papageorgis, 1961; Pfau & Burgoon, 1988; Pfau, Kenski, Nitz, & Sorenson, 1990). There has been relatively little work on the specific qualities that render atti­ tude changes resistant to attack. However, the existing data have supported the view that attitudes are more resistant to attack when they are accessible (Bassili & Fletcher, 1991) and have resulted from considerable issue-relevant elabora­ tion. For example, Haugtvedt and Petty (1992) provided subjects who were high or low in need for cognition with an initial message about the safety of a food additive. This initial message, containing strong arguments from an expert source, was followed by an opposite message containing rather weak arguments from a different expert source. Although both high and low need for cognition (NC) individuals were persuaded equally by the initial message, the attitudes of the high need for cognition subjects were more resistant to the attacking message. In addition, high NC individuals engaged in greater counterarguing of the attacking message. Similarly, Haugtvedt and Wegener (in press) found that subjects who encountered an initial message under conditions of high personal relevance (and thus processed it extensively) were relatively uninfluenced by a subsequent op­ posing message. In comparison, subjects who received the same messages under

116

PETTY, PRIESTER, WEGENER

conditions of low relevance were more influenced by the second communication. Furthermore, subjects in the high-relevance conditions engaged in more counterarguing of the opposing message. A strong initial attitude should prove especially resistant to a subsequent mes­ sage if that message is susceptible to counterarguing (i.e., presents a counterattitudinal position with weak or mixed arguments). Consistent with this reasoning, Wu and Shaffer (1987) varied attitude strength by manipulating whether the ini­ tial attitude toward a new consumer product was based on direct or indirect ex­ perience. They found that attitudes based on direct experience were more resistant to a counterattitudinal appeal, but more susceptible to a proattitudinal appeal than were attitudes based on indirect experience. Information gleaned from direct ex­ perience (e.g., the taste of the product) might be more accessible, held with greater confidence, and be linked more strongly to the attitude object allowing a person to more easily recognize the flaws in contrary information, but the merits in con­ gruent information. In addition, when initial attitudes were based on direct ex­ perience, attitudes were less influenced by the credibility of the source of the second message, regardless of whether the message was pro- or counterattitudi­ nal. That is, strong attitudes were less susceptible to peripheral cues. Attitude-Behavior Consistency

Perhaps the most important quality of attitudes for those interested in applica­ tions of persuasion theory (e.g., in politics, the marketplace, health, etc.) con­ cerns the ability of attitudes to predict people’s actions. A number of situational and dispositional factors have been shown to enhance the consistency of attitudes with behaviors. For example, attitudes are more predictive of behavior when: (a) the persons tested are of a certain personality type (e.g., are low in self­ monitoring, Snyder & Swann, 1976; or high in need for cognition, Cacioppo, Petty, Kao, & Rodriguez, 1986); (b) the attitudes in question are consistent with underlying beliefs (Norman, 1975); (c) the attitudes are based on high, rather than low, amounts of issue-relevant knowledge (e.g., Davidson et al., 1985; Kall­ gren & Wood, 1986); (d) the attitudes were likely formed as a result of issue­ relevant thinking (Cacioppo et al., 1986; Leippe & Elkin, 1987; Verplanken, 1991); and (e) cues in the situation indicate that the person’s attitude is relevant to the behavior (Borgida & Campbell, 1982). Still other factors have been noted as well (see Ajzen, 1988; Pieters, 1988, for reviews). A number of methodological considerations also have proved to be important if attitudes are to predict behaviors. In particular, the attitude and behavior should be measured at the same level of correspondence (e.g., general attitudes predict multi-act criteria, specific attitudes predict single behaviors; Ajzen & Fishbein, 1977; Bagozzi, 1981), and the attitude and behavioral measures should be as­ sessed close in time (Davidson & Jaccard, 1979). If people are asked to think about the basis of their attitudes just prior to attitude measurement, attitude-

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

117

behavior consistency could be reduced if thinking produces an expressed attitude that is not representative of the true one (Wilson et al., 1989). For example, if the central merits of an attitude object are affectively based, but a thinking task makes cognitive rather than affective information salient prior to attitude expres­ sion, the attitude expressed after thought will be less predictive of behavior than an attitude expressed without thought, especially if the behavior is affectively based as well (Millar & Tesser, 1986b; see Millar & Tesser, 1992, for a review). A related point is that attitudes predict behaviors better when the same attributes of the attitude object are salient at both the time of attitude measurement and the time of behavioral expression (Shavitt & Fazio, 1991). In general, behavioral prediction can be improved by considering attitudes toward alternative courses of action (e.g., Jaccard, Dittus, Radecki, & Wilson, in press; Jaccard, Helbig, Wan, Gutman, & Kritz-Silverstein, 1990) and by including predictor variables other than attitudes (e.g., societal and personal norms, Ajzen & Fishbein, 1977; habits, Ronis, Yates, & Kirscht, 1989; Triandis, 1977; and perceived control and self-efficacy, Azjen, 1991). The effect of attitudes on behavior is mediated by behavioral intentions (Ajzen & Fishbein, 1980), at least when the intentions are formed clearly and the behavior is costly (i.e., likely to generate behavior­ relevant thought; Bagozzi & Yi, 1989). For less consequential behaviors, the im­ pact of attitudes should be more direct (Fazio, 1990). Although much research has examined how methodological factors and exist­ ing characteristics of attitudes (e.g., extent of knowledge), people (e.g., person­ ality), and situations (e.g., time pressure) moderate attitude-behavior consistency, relatively few studies have examined whether different attitude formation or change processes are related to the ability of newly formed or changed attitudes to predict behavior. However, some research has shown that attitudes correlate to a greater extent with behavioral intentions when the attitudes were formed under high than under low personal relevance conditions (Petty et al., 1983; Sivacek & Crano, 1982). In a coherent program of research, Fazio and his colleagues examined attitudes that were formed as a result of direct or indirect experience with the object, and found that the former are more predictive of behavior (see Fazio & Zanna, 1981, for a review). A primary reason for this is that attitudes based on direct experience are more accessible, and thus more available to color percep­ tion of the attitude object and guide behavior (Fazio, 1990). One reason that atti­ tudes based on direct experience are more accessible is that direct experience typically might lead to greater thought relevant to the attitude object than passive exposure to a persuasive message (Petty & Cacioppo, 1986a). The notion that accessibility is a moderator of attitude-behavior relations is an important one, and has the potential for integrating a considerable number of research findings. For example, many of the variables found to moderate attitude-behavior consistency might be explained by the accessibility notion. Thus, low self-monitors have more accessible attitudes than high self-monitors (Kardes, Sanbonmatsu, Voss, & Fazio, 1986), and attitudes formed by the central route

118

PETTY, PRIESTER, WEGENER

and via direct experience are more accessible than attitudes formed by the peripher­ al route (Rennier, 1988) or those that are based only on indirect experience (Fa­ zio & Zanna, 1981).

SUMMARY AND CONCLUSIONS

Our goal in this chapter was to outline the major processes of attitude change. Over the past few decades, much has been learned about the underlying deter­ minants and consequences of different attitude change processes. We argued that it is useful to divide the theoretical processes responsible for modifying attitudes into those that emphasize effortful thinking about the central merits of the atti­ tude object and those that do not. This framework allows understanding and predic­ tion of what variables affect attitudes and in what general situations. It also permits understanding and prediction of the consequences of attitude change. We em­ phasized that all attitudes (whether toward oneself, other people, objects, or is­ sues) can be based on cognitive, affective, and/or behavioral information, and that any one variable can have an impact on persuasion by invoking different processes under specifiable conditions of elaboration likelihood. Also, we noted that attitudes that appear identical when measured can be quite different in their temporal persistence, resistance, or ability to predict behavior. Thus, in attempting to account for the great swings in public opinion that oc­ curred early in the 1992 presidential campaign, it is plausible that many of the initial shifts were the result of peripheral processes. At the beginning of a politi­ cal campaign, motivation to process is relatively low and attitudes toward the candidates (especially the challengers) are relatively weak. Because of this, atti­ tudes could be modified and changed quite easily. In fact, people’s responses to political polls at this point in the campaign simply might be constructed on the basis of whatever information is momentarily salient. However, as the campaign continues, involvement should increase, and people’s attitudes should become more crystalized as a result of enhanced information processing activity. By the end of the campaign, attitudes should become relatively stable, resistant to change, and predictive of voting behavior. Although much progress has been made in understanding the various process­ es responsible for attitude change, much work remains to be done. We hope that the next decade will bring advances in several areas. First, greater appreciation is needed for the view that any one variable is capable of multiple roles in the persuasion process. At present, most studies still focus on the “one” true process by which a variable has an impact on attitudes. Second, more attention should be paid to the various processes by which likelihood or truth judgments are made. Like evaluative judgments, it appears that likelihood judgments can result from relatively effortful or noneffortful processes. A third area of research that war­ rants increased attention concerns the consequences of attitudes changed by

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

119

different means. Although considerable work has examined the extent to which existing attitudes have properties associated with strong, rather than weak, atti­ tudes (e.g., are resistant to change and predict behavior), relatively little work has been conducted on the strength consequences of newly formed or changed attitudes. A fourth area that is likely to engage the interest of researchers con­ cerns the emotional bases of attitudes. Although important work on the cognitive foundations of attitudes and the cognitive structure of opinions undoubtedly will continue, the next decade will likely bring new ways to conceptualize and inves­ tigate the role of affect in persuasion. What roles can affect play in persuasion and what processes can it elicit? What are the consequences of affective versus cognitive versus behavioral persuasion? Under what circumstances and for which individuals is each more effective? Finally, research on attitudes is relevant to so much of the domain of social psychology. Yet, research in some relevant areas (e.g., prejudice, the self, group influence, etc.) has proceeded mostly in isolation from work on attitudes and viceversa. We believe that the potential for integrative linkage is great. For example, the finding in the “attitudes” domain that a positivity bias in social judgment (i.e., an attraction effect; Zajonc, 1968b) requires fewer cognitive resources than a balance effect (Cacioppo & Petty, 1981a) is similar to the finding in the “self’ domain that people prefer self-enhancing information when deprived of cogni­ tive resources, but prefer self-verifying (i.e., consistent) information otherwise (Swann, Hixon, Stein-Seroussi, & Gilbert, 1990). Also, the general idea high­ lighted in this chapter that people’s responses to persuasive communications can be based on careful and effortful analysis or more cue-based processes is similar to and compatible with emerging findings in nonpersuasion domains. For example, recent research suggests that people are more likely to use rela­ tively simple response strategies (e.g., choosing the first response that seems reasonable) when answering opinion surveys when motivation and ability to think are low than when they are high (Krosnick, 1991), and people are more likely to use activated stereotypes when ability and/or motivation to think about individu­ ating information is low (e.g., Bodenhausen, 1990; Fiske & Pavelchak, 1986; Gilbert & Hixon, 1991). We hope that in the next decade, these and other con­ ceptually related research streams will become more closely linked as investiga­ tors in each area of social psychology take advantage of the significant advances in the others.

ACKNOWLEDGMENTS

We are grateful to Bob Wyer, Tom Ostrom, and Bill McGuire for their helpful comments on a previous version of this chapter. Preparation of this chapter was facilitated by an NSF grant to REP (BNS 9021647) and NIMH traineeships to the second and third authors (T32 MH19728).

120

PETTY, PRIESTER, WEGENER

REFERENCES Abelson, R. P. (in press). Consequences of attitude extremity. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Abelson, R. P., Kinder, D. R., Peters, M. D., & Fiske, S. T. (1982). Affective and semantic com­ ponents in political perception. Journal o f Personality and Social Psychology, 42, 619-630. Abelson, R. P., & Prentice, D. A. (1989). Beliefs as possessions: A functional perspective. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Ed.), Attitude structure and junction (pp. 361-381). Hillsdale, NJ: Lawrence Erlbaum Associates. Ahlering, R. F., & Parker, L. D. (1989). Need for cognition as a moderator of the primacy effect. Journal o f Research in Personality, 23, 313-317. Ajzen, I. (1988). Attitudes, personality, and behavior. Chicago: Dorsey. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179-211. Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84, 888-918. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Engle­ wood Cliffs, NJ: Prentice-Hall. Alba, J. W ., & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal o f Consumer Research, 13, 411-454. Alba, J. W ., & Marmorstein, H. (1987). The effects of frequency of knowledge on consumer deci­ sion making. Journal o f Consumer Research, 14, 14-25. Anderson, C. A., Lepper, M. R., & Ross, L. (1980). Perseverance of social theories: The role of explanation in the persistence of discredited information. Journal o f Personality and Social P sy­ chology, 39, 1037-1049. Anderson, J. R. (1983). The architecture o f cognition. Cambridge, MA: Harvard University Press. Anderson, N. H. (1971). Integration theory and attitude change. Psychological Review, 78, 171-206. Anderson, N. H. (1981). Integration theory applied to cognitive responses and attitudes. In R. Petty, T. Ostrom, & T. Brock (Eds.), Cognitive responses in persuasion (pp. 361-397). Hillsdale, NJ: Lawrence Erlbaum Associates. Anderson, N. H., & Hubert, S. (1963). Effects of concomitant verbal recall on order effects in per­ sonality impression formation. Journal o f Verbal Learning and Verbal Behavior, 2, 379-391. Arkes, H. R., Boehm, L. E., & Xu, G. (1991). Determinants of judged validity. Journal o f Ex­ perimental Social Psychology, 27, 576-605. Aronson, E. (1968). Dissonance theory: Progress and problems. In R. P. Abelson, E. Aronson, W. J. McGuire, T. M. Newcomb, M. J. Rosenberg, & P. H. Tannenbaum (Eds.), Theories o f cognitive consistency: A sourcebook. Chicago: Rand McNally. Aronson, E. (1992). The return of the repressed: Dissonance theory makes a comeback. Psychologi­ cal Inquiry, 3, 303-311. Asch, S. E. (1956). Studies of independence and conformity: Vol. 1. A minority of one against a unanimous majority. Psychological Monographs, 70 (9, Whole No. 416). Axsom, D., Yates, S. M ., & Chaiken, S. (1987). Audience response as a heuristic cue in persua­ sion. Journal o f Personality and Social Psychology, 53, 30-40. Bagozzi, R. P. (1981). Attitudes, intentions, and behavior: A test of some key hypotheses. Journal o f Personality and Social Psychology, 41, 607-627. Bagozzi, R. P. (1984). Expectancy-value attitude models: An analysis of critical measurement is­ sues. International Journal o f Marketing, 1, 295-310. Bagozzi, R. P. (1985). Expectancy-value attitude models: An analysis of critical theoretical issues. International Journal o f Research in Marketing, 2, 43-60. Bagozzi, R. P., & Yi, Y. (1989). The degree of intention formation as a moderator of the attitude-behavior relation. Social Psychology Quarterly, 52, 266-279.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

121

Baker, S. M ., & Petty, R. E. (1993). M ajority and minority influence: Source advocacy as a deter­ minant o f m essage scrutiny. Unpublished manuscript, Ohio State University, Columbus. Bargh, J. A., Chaiken, S., Govender, R., & Pratto, F. (1992). The generality of the automatic atti­ tude activation effect. Journal o f Personality and Social P sychology, 62, 893-912. Bargh, J. A., & Thein, R. D. (1985). Individual construct accessibility, person memory, and the recall-judgment link: The case of information overload. Journal o f Personality and Social P sy­ chology, 49, 1129-1146. Bassili, J. N., & Fletcher, J. F. (1991). Response-time measurement in survey research: A method for CATI and a new look at nonattitudes. Public Opinion Quarterly, 55, 331-346. Batra, R., & Ray, M. L. (1986). Situational effects of advertising repetition: The moderating in­ fluence of motivation, ability, and opportunity to respond. Journal o f Consumer Research, 12, 432-445. Batra, R., & Stayman, D. M. (1990). The role of mood in advertising effectiveness. Journal o f Con­ sumer Research, 17, 203-214. Bern, D. J. (1965). An experimental analysis of self-persuasion. Journal o f Experimental Social Psy­ chology, 1, 199-218. Bern, D. J. (1970). Beliefs, attitudes, and human affairs. Belmont, CA: Brooks/Cole. Bern, D. J. (1972). Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 6, pp. 1-62). New York: Academic Press. Berkowitz, L., & Devine, P. G. (1989). Research traditions, analysis, and synthesis in social psy­ chological theories: The case of dissonance theory. Personality and Social Psychology Bulletin, 15, 493-507. Birnbaum, M. H., & Stegner, S. E. (1979). Source credibility in social judgment: Bias, expertise, and the judge’s point of view. Journal o f Personality and Social Psychology, 37, 48-74. Birnbaum, M. H., Wong, R., & Wong, L. K. (1976). Combining information from sources that vary in credibility. Memory and Cognition, 4, 330-336. Blaney, P. H. (1986). Affect and memory: A review. Psychological Bulletin, 99, 229-246. Blascovich, J., Ernst, J. M ., Tomaka, J., Kelsey, R. M ., Salomon, K. L., & Fazio, R. H. (1993). Attitude accessibility as a moderator of autonomic reactivity during decision making. Journal o f Personality and Social Psychology, 64, 165-176. Bless, H., Bohner, G., Schwarz, N., & Strack, F. (1990). Mood and persuasion: A cognitive response analysis. Personality and Social Psychology Bulletin, 16, 331-345. Bodenhausen, G. V. (1990). Stereotypes as judgmental heuristics: Evidence of circadian variations in discrimination. Psychological Science, 1, 319-322. Boninger, D., Berent, M ., Krosnick, J. A., & Fabrigar, L. (in press). The causes, correlates, and consequences of attitude importance. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Boninger, D. S., Brock, T. C., Cook, T. D., Gruder, C. L., & Romer, D. (1990). Discovery of reliable attitude change persistence resulting from a transmitter tuning set. Psychological Science, 1, 268-271. Borgida, E., & Campbell, B. (1982). Belief relevance and attitude-behavior consistency: The moderating role of personal experience. Journal o f Personality and Social Psychology, 42, 239-247. Bornstein, R. F. (1989). Exposure and affect: Overview and meta-analysis of research, 1968-1987. Psychological Bulletin, 106, 265-289. Bornstein, R. F., & D’Agostino, P. R. (1992). Stimulus recognition and the mere exposure effect. Journal o f Personality and Social Psychology, 63, 545-552. Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129-148. Bozzolo, A. (1993). Cognitive momentum analysis o f program context and media placem ent as de­ terminants o f advertising effectiveness. Unpublished doctoral dissertation, Ohio State University, Columbus, OH. Bozzolo, A. M ., & Brock, T. C. (1992). Unavailability effects on message processing: A theoretical analysis and an empirical test. Basic and A pplied Social Psychology, 13, 93-101.

122

PETTY, PRIESTER, WEGENER

Bray, R. M., Johnson, D., & Chilstron, J. T. (1982). Social influence by group members with minority opinions: A comparison of Hollander and Moscovici. Journal o f Personality and Social Psycholo­ gy, 43, 78-88. Breckler, S. J. (1984). Empirical validation of affect, behavior, and cognition as distinct components of attitude. Journal o f Personality and Social Psychology, 47, 1191-1205. Breckler, S. J. (in press). Emotion and attitude change. In M. Lewis & J. M. Haviland (Eds.), Handbook o f emotion. New York: Guilford. Breckler, S. J., & Wiggins, E. C. (1989a). Affect versus evaluation in the structure of attitudes. Journal o f Experimental Social Psychology, 25, 253-271. Breckler, S. J., & Wiggins, E. C. (1989b). On defining attitude and attitude theory: Once more with feeling. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 407-427). Hillsdale, NJ: Lawrence Erlbaum Associates. Breckler, S. J., & Wiggins, E. C. (1991). Cognitive responses in persuasion: Affective and evalua­ tive determinants. Journal o f Experimental Social Psychology, 27, 180-200. Brehm, J. W. (1966). A theory o f psychological reactance. San Diego, CA: Academic Press. Brock, T. C. (1967). Communication discrepancy and intent to persuade as determinants of counter­ argument production. Journal o f Experimental Social Psychology, 3, 296-309. Brock, T. C. (1968). Implications of commodity theory for value change. In A. G. Greenwald, T. C. Brock, & T. M. Ostrom (Eds.), P sychological foundations o f attitudes (pp. 243-296). New York: Academic Press. Bracks, M ., Armstrong, G. M ., & Goldberg, M. E. (1988). Children’s use of cognitive defenses against television advertising: A cognitive response approach. Journal o f Consumer Research, 14, 471-482. Burke, R. R., DeSarbo, W. S., Oliver, R. L., & Robertson, T. S. (1988). Deception by implication: An experimental investigation. Journal o f Consumer Research, 14, 482-494. Bumkrant, R. E., & Howard, D. J. (1984). Effects of the use of introductory rhetorical questions versus statements on information processing. Journal o f Personality and Social Psychology, 47, 1218-1230. Burnkrant, R. E., & Unnava, R. (1989). Self-referencing: A strategy for increasing processing of message content. Personality and Social Psychology Bulletin, 15, 628-638. Burnstein, E., & Vinokur, A. (1975). What a person thinks upon learning he has chosen differently from others: Nice evidence for the persuasive-arguments explanation of choice shifts. Journal o f Experimental Social Psychology, 112, 412-426. Burnstein, E., Vinokur, A., & Trope, Y. (1973). Interpersonal comparison versus persuasive ar­ gumentation: A more direct test of alternative explanations for group induced shifts in individual choice. Journal o f Experimental Social Psychology, 9, 236-245. Cacioppo, J. T ., Marshall-Goodell, B. S., Tassinary, L. G., & Petty, R. E. (1992). Rudimentary determinants of attitudes: Classical conditioning is more effective when prior knowledge about the attitude stimulus is low than high. Journal o f Experimental Social Psychology, 28, 207-233. Cacioppo, J. T., & Petty, R. E. (1979a). Attitudes and cognitive response: An electro-physiological approach. Journal o f Personality and Social Psychology, 37, 2181-2199. Cacioppo, J. T., & Petty, R. E. (1979b). Effects of message repetition and position on cognitive responses, recall, and persuasion. Journal o f Personality and Social Psychology, 37, 97-109. Cacioppo, J. T ., & Petty, R. E. (1981a). Effects of extent of thought on the pleasantness ratings of P-O-X triads: Evidence for three judgmental tendencies in evaluating social situations. Journal o f Personality and Social Psychology, 40, 1000-1009. Cacioppo, J. T., & Petty, R. E. (1981b). Social psychological procedures for cognitive response assessment: The thought listing technique. In T. Merluzzi, C. Glass, & M. Genest (Eds.), Cogni­ tive assessment (pp. 309-342). New York: Guilford. Cacioppo, J. T ., & Petty, R. E. (1982). The need for cognition. Journal o f Personality and Social Psychology, 42, 116-131.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

123

Cacioppo, J. T ., & Petty, R. E. (1984). The need for cognition: Relationship to attitudinal process­ es. In R. McGlynn, J. Maddux, C. Stoltenberg, & J. Harvey (Eds.), Social perception in clinical and counseling psychology (pp. 113-119). Lubbock, TX: Texas Tech Press. Cacioppo, J. T ., & Petty, R. E. (1985). Central and peripheral routes to persuasion: The role of message repetition. In A. Mitchell & L. Alwitt (Eds.), Psychological processes and advertising effects (pp. 91-111). Hillsdale, NJ: Lawrence Erlbaum Associates. Cacioppo, J. T., & Petty, R. E. (1989). Effects of message repetition on argument processing, recall, and persuasion. Basic and A pplied Social Psychology, 10, 3-12. Cacioppo, J. T., Petty, R. E., & Berntson, G. G. (1991). Persuasion. In R. Dulbecco (Ed.), Ency­ clopedia o f human biology (pp. 799-810). Orlando, FL: Academic Press. Cacioppo, J. T ., Petty, R. E., Kao, C. F., & Rodriguez, R. (1986). Central and peripheral routes to persuasion: An individual difference perspective. Journal o f Personality and Social Psycholo­ gy, 51, 1032-1043. Cacioppo, J. T ., Petty, R. E., & Morris, K. J. (1983). Effects of need for cognition on message evaluation, recall, and persuasion. Journal o f Personality and Social Psychology, 45, 805-818. Cacioppo, J. T ., Petty, R. E., & Sidera, J. (1982). The effects of salient self-schema on the evalua­ tion of proattitudinal editorials: Top-down versus bottom-up message processing. Journal o f Ex­ perimental Social Psychology, 18, 324-338. Cacioppo, J. T., Priester, J. R., & Berntson, G. G. (1993). Rudimentary determinants of attitudes II: Arm flexion and extension have differential effects on attitudes. Journal o f Personality and Social Psychology, 65, 5-17. Chaiken, S. (1980). Heuristic versus systematic information processing in the use of source versus message cues in persuasion. Journal o f Personality and Social Psychology, 39, 752-766. Chaiken, S. (1987). The heuristic model of persuasion. In M. P. Zanna, J. M. Olson, & C. P. Her­ man (Eds.), Social influence: The Ontario Symposium (Vol. 5, pp. 3-39). Hillsdale, NJ: Lawrence Erlbaum Associates. Chaiken, S., & Baldwin, M. W. (1981). Affective-cognitive consistency and the effect of salient behavioral information on the self-perception of attitudes. Journal o f Personality and Social P sy­ chology, 41, 1-12. Chaiken, S., & Eagly, A. H. (1976). Communication modality as a determinant of message persua­ siveness and message comprehensibility. Journal o f Personality and Social Psychology, 34, 605-614. Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and systematic processing within and beyond the persuasion context. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 212-252). New York: Guilford. Chaiken, S., & Maheswaran, D. (in press). Heuristic processing can bias systematic processing: Ef­ fects of source credibility, argument ambiguity, and task importance on attitude judgment. Jour­ nal o f Personality and Social Psychology. Chaiken, S ., Pomerantz, E. M ., & Giner-Sorolla, R. (in press). Structural consistency and attitude strength. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences.

Hillsdale, NJ: Lawrence Erlbaum Associates. Chaiken, S., & Yates, S. M. (1985). Affective-cognitive consistency and thought-induced attitude polarization. Journal o f Personality and Social Psychology, 49, 1470-1481. Chattopadhyay, A., & Alba, J. W. (1988). The situational importance of recall and inference in con­ sumer decision making. Journal o f Consumer Research, 15, 1-12. Chen, H. C., Reardon, R., Rea, C ., & Moore, D. J. (1992). Forewarning of content and involve­ ment: Consequences for persuasion and resistance to persuasion. Journal o f Experimental Social Psychology, 28, 523-541. Cialdini, R. B. (1987). Compliance principles of compliance professionals: Psychologists of necessi­ ty. In M. P. Zanna, J. M. Olson, & C. P. Herman (Eds.), Social influence: The Ontario Symposi­ um (Vol. 5, pp. 165-184). Hillsdale, NJ: Lawrence Erlbaum Associates. Clark, M. S., & Isen, A. M. (1982). Feeling states and social behavior. In A. Hastorf & A. M. Isen (Eds.), Cognitive social psychology (pp. 73-108). New York: Elsevier/North-Holland.

124

PETTY, PRIESTER, WEGENER

Cook, T. D., & Flay, B. R. (1978). The persistence of experimentally induced attitude change. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 11, pp. 1-57). New York: Academic Press. Cook, T. D ., Grader, C. L., Hennigan, K. M ., & Flay, B. R. (1979). History of the sleeper effect: Some logical pitfalls in accepting the null hypothesis. Psychological Bulletin, 86, 662-679. Cooper, J., & Fazio, R. H. (1984). A new look at dissonance theory. In L. Berkowitz (Ed.), Ad­ vances in experimental social psychology (Vol. 17, pp. 229-266). New York: Academic Press. Cooper, J., & Worchel, S. (1970). Role of undesired consequences in arousing cognitive dissonance. Journal o f Personality and Social Psychology, 16, 199-206. Cox, D. S., & Cox, A. D. (1988). What does familiarity breed: Complexity as a moderator of repe­ tition effects in advertisement evaluation. Journal o f Consumer Research, 15, 111-116. Crano, W. D. (in press). Attitude strength and vested interest. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Crano, W. D., Gorenflo, D. W ., & Shackelford, S. L. (1988). Overjustification, assumed consen­ sus, and attitude change: Further investigation of the incentive-aroused ambivalence hypothesis. Journal o f Personality and Social Psychology, 55, 12-22. Crites, S., Fabrigar, L., & Petty, R. (1993). Measuring the affective and cognitive properties o f atti­ tudes: Conceptual and m ethodological issues. Unpublished manuscript, Ohio State University, Columbus. Crocker, J., Fiske, S. T., & Taylor, S. E. (1984). Schematic bases of belief change. In R. Eiser (Ed.), Attitudinal judgm ent (pp. 197-226). New York: Springer-Verlag. Cronen, V. E., & Conville, R. L. (1975). Fishbein’s conception of belief strength: A theoretical, methodological, and experimental critique. Speech Monographs, 42, 143-150. Davidson, A. R. (in press). From attitudes to actions to attitude change. The effects of amount and accuracy of information. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequence. Hillsdale, NJ: Lawrence Erlbaum Associates. Davidson, A. R., & Jaccard, J. J. (1979). Variables that moderate the attitude-behavior relation: Results of a longitudinal survey. Journal o f Personality and Social Psychology, 37, 1364-1376. Davidson, A. R., Yantis, S., Norwood, M ., & Montano, D. E. (1985). Amount of information about the attitude object and attitude-behavior consistency. Journal o f Personality and Social Psycholo­ gy, 49, 1184-1198. DeBono, K. G. (1987). Investigating the social-adjustive and value-expressive functions of attitudes: Implications for persuasion processes. Journal o f Personality and Social Psychology, 52, 279-287. DeBono, K. G., & Harnish, R. J. (1988). Source expertise, source attractiveness, and the processing of persuasive information: A functional approach. Journal o f Personality and Social Psychology, 55, 541-546. DeBono, K. G., & Klein, C. (1993). Source expertise and persuasion: The moderating role of recipient dogmatism. Personality and Social Psychology Bulletin, 19, 167-173. Deci, E. L. (1975). Intrinsic motivation. New York: Plenum. Devine, P. G., & Ostrom, T. M. (1988). Dimensional versus information-processing approaches to social knowledge: The case of inconsistency management. In D. Bar-Tal & A. W. Kraglanski (Eds.), The social psychology o f knowledge (pp. 231-261). Cambridge: Cambridge Universi­ ty Press. Dillehay, R. C., Insko, C. A., & Smith, M. B. (1966). Logical consistency and attitude change. Journal o f Personality and Social Psychology, 3, 646-654. Eagly, A. H. (1974). Comprehensibility of persuasive arguments as a determinant of opinion change. Journal o f Personality and Social Psychology, 29, 758-773. Eagly, A. H ., & Chaiken, S. (1984). Cognitive theories of persuasion. In L. Berkowitz (Ed.), Ad­ vances in experimental social psychology (Vol. 17, pp. 268-361). New York: Academic Press. Eagly, A. H., & Chaiken, S. (1993). The psychology o f attitudes. Fort Worth, TX: Harcourt, Brace, Jovanovich.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

125

Eagly, A. H., & Chaiken, S. (in press). Attitude strength, attitude structure, and resistance to change. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hills­ dale, NJ: Lawrence Erlbaum Associates. Eagly, A. H., Chaiken, S., & Wood, W. (1981). An attribution analysis of persuasion. In J. H. Harvey, W. J. Ickes, & R. F. Kidd (Eds.), New direction in attribution research (Vol. 3, pp. 37-62). Hillsdale, NJ: Lawrence Erlbaum Associates. Eagly, A. H ., & Telaak, K. (1972). Width of the latitude of acceptance as a determinant of attitude change. Journal o f Personality and Social Psychology, 23, 388-397. Eagly, A. H., & Warren, R. (1976). Intelligence, comprehension, and opinion change. Journal o f Personality, 44, 226-242. Eagly, A. H., Wood, W ., & Chaiken, S. (1978). Causal inferences about communicators and their effect on opinion change. Journal o f Personality and Social Psychology, 36, 424-435. Eaves, L. J., Eysenck, H. J., & Martin, N. G. (1989). Genes, culture, and personality: An empirical approach. London: Academic Press. Edwards, K. (1990). The interplay of affect and cognition in attitude formation and change. Journal o f Personality and Social Psychology, 59, 202-216. Elkin, R. A., & Leippe, M. R. (1986). Physiological arousal, dissonance, and attitude change: Evi­ dence for a dissonance-arousal link and a “Don’t remind me” effect. Journal o f Personality and Social Psychology, 51, 55-65. Elms, A. C. (1966). Influence of fantasy ability on attitude change through role-playing. Journal o f Personality and Social Psychology, 4, 36-43. Erber, R. (1991). Affective and semantic priming: Effects of mood on category accessibility and inference. Journal o f Experimental Social Psychology, 27, 480-498. Evans, M. G. (1991). The problem of analyzing multivariate composites: Interactions revisited. Ameri­ can Psychologist, 46, 6-15. Fazio, R. H. (1986). How do attitudes guide behavior? In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook o f motivation and cognition: Foundations o f social behavior (pp. 204-243). New York: Guilford. Fazio, R. H. (1990). Multiple processes by which attitudes guide behavior: The MODE model as an integrative framework. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 75-109). San Diego, CA: Academic Press. Fazio, R. H. (in press). Attitudes as object-evaluation associations: Determinants, consequences, and correlates of attitude accessibility. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: An­ tecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Fazio, R. H., Blascovich, J., & Driscoll, D. M. (1992). On the functional value of attitudes: The influence of accessible attitudes upon the ease and quality of decision-making. Personality and Social Psychology Bulletin, 18, 388-401. Fazio, R. H., & Cooper, J. (1983). Arousal in the dissonance process. In J. T. Cacioppo & R. E. Petty (Eds.), Social psychophysiology: A sourcebook (pp. 122-152). New York: Guilford. Fazio, R. H., Herr, P. M ., & Olney, T. J. (1984). Attitude accessibility following a self-perception process. Journal o f Personality and Social Psychology, 47, 277-286. Fazio, R. H., Sanbonmatsu, D. M ., Powell, M. C., & Kardes, F. R. (1986). On the automatic acti­ vation of attitudes. Journal o f Personality and Social Psychology, 50, 229-238. Fazio, R. H., & Williams, C. J. (1986). Attitude accessibility as a moderator of the attitude-perception and attitude-behavior relations: An investigation of the 1984 presidential election. Journal o f Per­ sonality and Social Psychology, 51, 505-514. Fazio, R. H., & Zanna, M. P. (1978). Attitudinal qualities relating to the strength of the attitude-behavior relationship. Journal o f Experimental Social Psychology, 14, 398-408. Fazio, R. H., & Zanna, M. P. (1981). Direct experience and attitude behavior consistency. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 14, pp. 161-202). New York: Academic Press.

126

PETTY, PRIESTER, WEGENER

Fazio, R. H., Zanna, M. P., & Cooper, J. (1977). Dissonance and self-perception: An integrative view of each theory’s proper domain of application. Journal o f Experimental Social Psychology, 13, 464-419.

Feldman, J. M ., & Lynch, J. G., Jr. (1988). Self-generated validity and other effects of measure­ ment on belief, attitude, intention, and behavior. Journal o f A pplied Psychology, 73, 421-435. Festinger, L. (1957). A theory o f cognitive dissonance. Evanston, IL: Row, Peterson. Festinger, L. (1964). Conflict, decision, and dissonance. Stanford, CA: Stanford University Press. Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal o f Abnormal and Social Psychology, 58, 203-210. Fishbein, M. (1980). A theory of reasoned action: Some applications and implications. In H. E. Howe, Jr. & M. M. Page (Eds.), Nebraska symposium on motivation, 1979 (Vol. 27, pp. 65-116). Lin­ coln, NE: University of Nebraska Press. Fishbein, M ., & Ajzen, I. (1975). Belief, attitude, intention, and behavior. Reading, MA: AddisonWesley. Fishbein, M ., & Ajzen, I. (1981). Acceptance, yielding and impact: Cognitive processes in persua­ sion. In R. E. Petty, T. M. Ostrom, & T. C. Brock (Eds.), Cognitive responses in persuasion (pp. 339-359). Hillsdale, NJ: Lawrence Erlbaum Associates. Fiske, S. T., & Pavelchak, M. A. (1986). Category-based versus piecemeal-based affective responses: Developments in schema-triggered affect. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook o f motivation and cognition: Foundations o f social behavior (pp. 167-203). New York: Guilford. Frey, D. (1986). Recent research on selective exposure to information. Advances in Experimental Social Psychology, 19, 41-80. Gardner, M. P. (1985). Mood states and consumer behavior: A critical review. Journal o f Consumer Research, 12, 281-300. Gerard, H. B., Blevans, S. A., & Malcolm, T. (1964). Self-evaluation and the evaluation of choice alternatives. Journal o f Personality, 32, 395-410. Gilbert, D. T. (1991). How mental systems believe. American Psychologist, 46, 107-119. Gilbert, D. T., & Hixon, J. G. (1991). The trouble of thinking: Activation and application of stereo­ typic beliefs. Journal o f Personality and Social Psychology, 60, 509-517. Gilbert, D. T ., Krull, D. S., & Malone, P. S. (1990). Unbelieving the unbelievable: Some problems in the rejection of false information. Journal o f Personality and Social Psychology, 59, 601-613. Gilbert, D. T., Tafarodi, R. W ., & Malone, P. S. (1993). You can’t not believe everything you read. Journal o f Personality and Social Psychology, 65, 221-233. Gillig, P. M ., & Greenwald, A. G. (1974). Is it time to lay the sleeper effect to rest? Journal o f Personality and Social Psychology, 29, 132-139. Gleicher, F., & Petty, R. E. (1992). Expectations of reassurance influence the nature of fear-stimulated attitude change. Journal o f Experimental Social Psychology, 28, 86-100. Goldberg, M. E., & Gorn, G. J. (1987). Happy and sad TV programs: How they affect reactions to commercials. Journal o f Consumer Research, 14, 387-403. Gorn, G. J. (1982). The effects of music in advertising on choice behavior: A classical conditioning approach. Journal o f Marketing, 46, 94-101. Gorn, G. J., & Goldberg, M. E. (1980). Children’s responses to repetitive television commercials. Journal o f Consumer Research, 6, 421-424. Gouaux, C. (1971). Induced affective states and interpersonal attraction. Journal o f Personality and Social Psychology, 20, 687-695. Greenwald, A. G. (1968). Cognitive learning, cognitive response to persuasion, and attitude change. In A. G. Greenwald, T. C. Brock, & T. M. Ostrom (Eds.), Psychological foundations o f attitudes (pp. 147-170). New York: Academic Press. Greenwald, A. G. (1975). On the inconclusiveness of “crucial” cognitive tests of dissonance versus self-perception theory. Journal o f Experimental Social Psychology, 11, 490-499.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

127

Greenwald, A. G., & Albert, R. D. (1968). Acceptance and recall of improvised arguments. Journal o f Personality and Social Psychology, 8 , 31-34. Greenwald, A. G., & Leavitt, C. (1984). Audience involvement in advertising: Four levels. Journal o f Consumer Research, 11, 581-592. Greenwald, A. G., & Ronis, D. L. (1978). Twenty years of cognitive dissonance: Case study of a theory. Psychological Review, 85, 53-57. Gresham, L. G., & Shimp, T. A. (1985). Attitude toward the advertisement and brand attitude: A classical conditioning perspective. Journal o f Advertising, 14, 10-17. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and seman­ tics (Vol. 3, pp. 41-58). New York: Academic Press. Griffitt, W. B. (1970). Environmental effects on interpersonal affective behavior: Ambient effective temperature and attraction. Journal o f Personality and Social Psychology, 15, 240-244. Gross, S., Holtz, R., & Miller, N. (in press). Attitude certainty. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum As­ sociates. Gruenfeld, D. H., & Wyer, R. S., Jr. (1992). Semantics and pragmatics of social influence: How affirmations and denials affect beliefs in referent propositions. Journal o f Personality and Social Psychology, 62, 38-49. Grush, J. E. (1976). Attitude formation and mere exposure phenomena: A nonartifactual explanation of empirical findings. Journal o f Personality and Social Psychology, 33, 281-290. Hannah, D., & Sternthal, B. (1984). Detecting and explaining the sleeper effect. Journal o f Con­ sumer Research, 11, 632-642. Harkins, S. G., & Petty, R. E. (1981). Effects of source magnification of cognitive effort on atti­ tudes: An information processing view. Journal o f Personality and Social Psychology, 40, 401-413. Harkins, S. G., & Petty, R. E. (1987). Information utility and the multiple source effect. Journal o f Personality and Social Psychology, 52, 260-268. Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the judgment task is memory-based or on-line. Psychological Review, 93, 258-268. Haugtvedt, C. P., & Petty, R. E. (1992). Personality and persuasion: Need for cognition moderates the persistence and resistance of attitude changes. Journal o f Personality and Social Psychology, 63, 308-319. Haugtvedt, C. P., Petty, R. E., & Cacioppo, J. T. (1992). Need for cognition and advertising: Un­ derstanding the role of personality variables in consumer behavior. Journal o f Consumer Psychol­ ogy, 1, 239-260. Haugtvedt, C. P., & Wegener, D. T. (in press). Message order effects in persuasion: An attitude strength perspective. Journal o f Consumer Research. Haugtvedt, C. P., & Wegener, D. T. (1993, May). N eed fo r cognition and message order effects in persuasion. Paper presented at the annual meeting of the Midwestern Psychological Associa­ tion, Chicago. Hawkins, S. A., & Hoch, S. J. (1992). Low-involvement learning: Memory without evaluation. Journal o f Consumer Research, 19, 212-225. Heesacker, M. H., Petty, R. E., & Cacioppo, J. T. (1983). Field dependence and attitude change: Source credibility can alter persuasion by affecting message-relevant thinking. Journal o f Person­ ality, 51, 653-666. Heider, F. (1958). The psychology o f interpersonal relations. New York: Wiley. Henninger, M ., & Wyer, R. S. (1976). The recognition and elimination of inconsistencies among syllogistically related beliefs: Some new light on the “Socratic effect.” Journal o f Personality and Social Psychology, 34, 680-693. Herek, G. M. (1987). Can functions be measured? A new perspective on the functional approach to attitudes. Social Psychology Quarterly, 50, 285-303.

128

PETTY, PRIESTER, WEGENER

Higgins, E. T. (1989). Knowledge accessibility and activation: Subjectivity and suffering from unconscious sources. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 75-123). New York: Guilford Press. Higgins, E. T ., & McCann, C. D. (1984). Social encoding and subsequent attitudes, impressions, and memory: “Context-driven” and motivational aspects of processing. Journal o f Personality and Social Psychology, 47, 26-39. Himmelfarb, S. (1972). Integration and attribution theories in personality impression formation. Journal o f Personality and Social Psychology, 23, 309-313. Houston, D. A., & Fazio, R. H. (1989). Biased processing as a function of attitude accessibility: Making objective judgments subjectively. Social Cognition, 7, 51-66. Hovland, C. I. (1959). Reconciling conflicting results derived from experimental and survey studies of attitude change. American Psychologist, 14, 8-17. Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion: Psychological studies o f opinion change. New Haven, CT: Yale University Press. Hovland, C. I., Lumsdaine, A. A., & Sheffield, F. D. (1949). Experiments on mass communication. Princeton, NJ: Princeton University Press. Howard, D. J. (1990). Rhetorical question effects on message processing and persuasion: The role of information availability and the elicitation of judgment. Journal o f Experimental Social P sy­ chology, 26, 217-239. Howard-Pitney, B., Borgida, E ., & Omoto, A. M. (1986). Personal involvement: An examination of processing differences. Social Cognition, 4, 899-910. Huesmann, L. R., Eron, L. D., Klein, R., Brice, P., & Fischer, P. (1983). Mitigating the imitation of aggressive behaviors by changing children’s attitudes about media violence. Journal o f Person­ ality and Social Psychology, 44, 899-910. Hutton, D. G., & Baumeister, R. F. (1992). Self-awareness and attitude-change: Seeing oneself on the central route to persuasion. Personality and Social Psychology Bulletin, 18, 68-75. Ickes, W ., Robertson, E., Tooke, W ., & Teng, G. (1986). Naturalistic social cognitions: Methodology, assessment, and validation. Journal o f Personality and Social Psychology, 51, 66-82. Insko, C. A. (1967). Theories o f attitude change. New York: Appleton-Century-Crofts. Insko, C. A. (1981). Balance theory and phenomenology. In R. E. Petty, T. M. Ostrom, & T. C. Brock (Eds.), Cognitive responses in persuasion (pp. 309-338). Hillsdale, NJ: Lawrence Erlbaum Associates. Insko, C. A. (1984). Balance theory, the Jordan paradigm, and the Wiest tetrahedron. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 18, pp. 89-140). San Diego, CA: Academic Press. Insko, C. A., Turnbull, W ., & Yandell, B. (1974). Facilitative and inhibiting effects of distraction on attitude change. Sociometry, 37, 508-528. Isen, A. M ., Daubman, K. A., & Nowicki, G. P. (1987). Positive affect facilitates creative problem solving. Journal o f Personality and Social Psychology, 52, 1122-1131. Isen, A. M ., & Simmonds, S. (1978). The effect of feeling good on a helping task that is incompatible with good mood. Social Psychology Quarterly, 41, 346-349. Iyengar, S., Kinder, D. R., Peters, M. D., & Krosnick, J. A. (1984). The evening news and presidential evaluations. Journal o f Personality and Social Psychology, 46, 778-787. Jaccard, J., Dittus, P., Radecki, C., & Wilson, T. (in press). The construct of belief importance in attitude research. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Jaccard, J., Helbig, D. W ., Wan, C. K., Gutman, M ., & Kritz-Silverstein, D. S. (1990). Individual differences in attitude-behavior consistency: The prediction of contraceptive behavior. Journal o f A pplied Social Psychology, 20, 575-617. Jamieson, D. W ., & Zanna, M. P. (1989). Need for structure in attitude formation and expression. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 383-406). Hillsdale, NJ: Lawrence Erlbaum Associates.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

129

Janis, I. L., & King, B. T. (1954). The influence of role playing on opinion change. Journal o f Ab­ normal and Social Psychology, 49, 211-218. Jepson, C ., & Chaiken, S. (1990). Chronic issue-specific fear inhibits systematic processing of persuasive communications. Journal o f Social Behavior and Personality, 5, 61-84. Johnson, B. T ., & Eagly, A. H. (1989). Effects of involvement on persuasion: A meta-analysis. Psychological Bulletin, 106, 290-314. Johnson, B. T ., & Turco, R. M. (1993). Effects o f argument quality on persuasion as m oderated by knowledge and involvement: A meta-analysis. Unpublished manuscript, Syracuse University, Syracuse, New York. Johnson, E., & Tversky, A. (1983). Affect, generalization, and the perception of risk. Journal o f Personality and Social Psychology, 45, 20-31. Johnson, H. H.,*& Watkins, T. A. (1971). The effects of message repetition on immediate and delayed attitude change. Psychonomic Science, 22, 101-103. Jones, E. E., & Davis, K. E. (1965). From acts to dispositions: The attribution process in person perception. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2, pp. 219-266). New York: Academic Press. Jones, E. E., Rhodewalt, F., Berglas, S., & Skelton, J. A. (1981). Effects of strategic self-presentation on subsequent self-esteem. Journal o f Personality and Social Psychology, 41, 407-421. Judd, C. M ., & Brauer, M. (in press). Repetition and evaluative extremity. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Judd, C. M ., Drake, R. A., Downing, J. W ., & Krosnick, J. A. (1991). Some dynamic properties of attitude structures: Context-induced response facilitation and polarization. Journal o f Personal­ ity and Social Psychology, 60, 193-202. Judd, C. M ., & Krosnick, J. A. (1989). The structural bases of consistency among political attitudes: Effects of political expertise and attitude importance. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and junction (pp. 99-128). Hillsdale, NJ: Lawrence Erlbaum Associates. Judd, C. M ., & Kulik, J. A. (1980). Schematic effects of social attitudes on information processing and recall. Journal o f Personality and Social Psychology, 38, 569-578. Judd, C. M:, & Lusk, C. M. (1984). Knowledge structures and evaluative judgments: Effects of structural variables on judgment extremity. Journal o f Personality and Social Psychology, 46, 1193-1207. Kallgren, C. A., & Wood, W. (1986). Access to attitude-relevant information in memory as a determinant of attitude-behavior consistency. Journal o f Experimental Social Psychology, 22, 328-338. Kaplan, K. J. (1972). On the ambivalence-indifference problem in attitude theory and measurement: A suggested modification of the semantic differential technique. Psychological Bulletin, 77, 361-372. Kardes, F. R., Sanbonmatsu, D. M ., Voss, R. T., & Fazio, R. H. (1986). Self-monitoring and attitude accessibility. P ersonality and Social Psychology Bulletin, 12, 468-474. Kassin, S. M ., Reddy, M. E., & Tulloch, W. F. (1990). Juror interpretations of ambiguous evidence: The need for cognition, presentation order, and persuasion. Law and Human Behavior, 14, 43-55. Katz, D. (1960). The functional approach to the study of attitudes. Public Opinion Quarterly, 24, 163-204. Katz, D., & Stotland, E. (1959). A preliminary statement to a theory of attitude structure and change. In S. Koch (Ed.), Psychology: A study o f science (Vol. 3, pp. 423-475). New York: McGraw-Hill. Kelley, H. H. (1967). Attribution theory in social psychology. In D. Levine (Ed.), Nebraska sympo­ sium on motivation (Vol. 15, pp. 192-238). Lincoln, NE: University of Nebraska Press. Kelman, H. C. (1958). Compliance, identification, and internalization: Three processes of attitude change. Journal o f Conflict Resolution, 2, 51-60.

130

PETTY, PRIESTER, WEGENER

Kelman, H. C., & Hovland, C. I. (1953). “Reinstatement” of the communicator in delayed measurement of opinion change. Journal o f Abnormal and Social Psychology, 48, 327-335. Kiesler, C. A. (1971). The psychology o f commitment: Experiments linking behavior to beliefs. New York: Academic Press. Kiesler, C. A., Collins, B. E., & Miller, N. (1969). Attitude change: A critical analysis o f theoreti­ cal applications. New York: Wiley. Kisielius, J., & Stemthal, B. (1984). Detecting and explaining vividness effects in attitudinal judgments. Journal o f Marketing Research, 21, 54-64. Kisielius, J., & Stemthal, B. (1986). Examining the vividness controversy: An availability-valence interpretation. Journal o f Consumer Research, 12, 418-431. Krosnick, J. A. (1988). Attitude importance and attitude change. Journal o f Experimental Social Psy­ chology, 24, 240-255. Krosnick, J. A. (1991). Response strategies for coping with the cognitive demands of attitude measures in surveys. Applied Cognitive Psychology, 5, 213-236. Krosnick, J. A., & Abelson, R. P. (1992). The case for measuring attitude strength in surveys. In J. M. Tanur (Ed.), Questions about questions: Inquiries into the cognitive bases o f surveys (pp. 177-203). New York: Russell Sage Foundation. Krosnick, J. A., Betz, A. L., Jussim, L. J., & Lynn, A. R. (1992). Subliminal conditioning of attitudes. Personality and Social Psychology Bulletin, 18, 152-162. Krosnick, J. A., Boninger, D. S., Chuang, Y. C ., Berent, M ., & Carnot, C. G. (in press). Atti­ tude strength: One construct or many related constructs? Journal o f Personality and Social Psychology.

Kruglanski, A. W ., & Freund, T. (1983). The freezing and unfreezing of lay-inferences: Effects on impressional primacy, ethnic stereotyping, and numerical anchoring. Journal o f Experimental Social Psychology, 19, 448-468. Kruglanski, A. W ., & Mackie, D. M. (1990). Majority and minority influence: A judgmental process analysis. In W. Stroebe & M. Hewstone (Eds.), European review o f social psychology (Vol. 1, pp. 229-262). New York: Wiley. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480-498. Kunst-Wilson, W. R., & Zajonc, R. B. (1980). Affective discrimination of stimuli that cannot be recognized. Science, 207, 557-558. Kuykendall, D., & Keating, J. P. (1990). Mood and persuasion. Psychology and Marketing, 7, 1-9. Leippe, M. R., & Elkin, R. A. (1987). When motives clash: Issue involvement and response involvement as determinants of persuasion. Journal o f Personality and Social Psychology, 52, 269-278. Lepper, M. R., Greene, D., & Nisbett, R. E. (1973). Undermining children’s intrinsic interest with extrinsic reward: A test of the “overjustification” hypothesis. Journal o f Personality and Social Psychology, 28, 129-137. Lepper, M. R., Ross, L., & Lau, R. R. (1986). Persistence of inaccurate beliefs about the self: Perseverance effects in the classroom. Journal o f Personality and Social Psychology, 50, 482-491. Levine, J. M ., Saxe, L ., & Harris, H. J. (1976). Reaction to opinion deviance: Impact of deviate’s direction and distance of movement. Sociometry, 39, 97-107. Lewan, P. C., & Stotland, E. (1961). The effects of prior information on susceptibility to an emotional appeal. Journal o f Abnormal and Social Psychology, 62, 450-453. Liberman, A., & Chaiken, S. (1991). Value conflict and thought-induced attitude change. Journal o f Experimental Social Psychology, 27, 203-216. Liberman, A., & Chaiken, S. (1992). Defensive processing of personally relevant health messages. Personality and Social Psychology Bulletin, 18, 669-679.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

131

Lichtenstein, M ., & Srull, T. K. (1985). Conceptual and methodological issues in examining the relationship between consumer memory and judgment. In L. F. Alwitt & A. A. Mitchell (Eds.), Psychological processes and advertising effects: Theory, research, and application (pp. 113-128). Hillsdale, NJ: Lawrence Erlbaum Associates. Lingle, J. H ., & Ostrom, T. M. (1981). Principles of memory and cognition in attitude formation. In R. E. Petty, T. M. Ostrom, & T. C. Brock (Eds.), Cognitive responses in persuasion (pp. 399-420). Hillsdale, NJ: Lawrence Erlbaum Associates. Linville, P. W. (1982). The complexity-extremity effect and age-based stereotyping. Journal o f P er­ sonality and Social Psychology, 42, 193-211. Lord, C. G., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude polarization: The effects of prior theories on subsequently considered evidence. Journal o f Personality and Social Psychology, 37, 2098-2109. Losch, M. E., & Cacioppo, J. T. (1990). Cognitive dissonance may enhance sympathetic tonus, but attitudes are changed to reduce negative affect rather than arousal. Journal o f Experimental Social Psychology, 26, 289-304. Love, R. E ., & Greenwald, A. G. (1978). Cognitive responses to persuasion as mediators of opinion change. Journal o f Social Psychology, 104, 231-241. Lutz, R. J. (1975). Changing brand attitudes through modification of cognitive structure. Journal o f Consumer Research, 1, 49-59. Lutz, R. J. (1977). An experimental investigation of causal relations among cognitions, affect, and behavioral intention. Journal o f Consumer Research, 3, 197-208. Lydon, J., Zanna, M. P., & Ross, M. (1988). Bolstering attitudes by autographical recall: Attitude persistence and selective memory. Personality and Social Psychology Bulletin, 14, 78-86. Maass, A., & Clark, R. D. (1983). Internalization versus compliance: Differential processes underlying minority influence and conformity. European Journal o f Social Psychology, 13, 197-215. Maass, A., & Clark, R. D. (1984). Hidden impact of minorities: Fifteen years of minority influence research. P sychological Bulletin, 95, 428-450. Maass, A., Clark, R. D ., & Haberkorn, G. (1982). The effects of differential ascribed category membership and norms on minority influence. European Journal o f Social Psychology, 12, 89-104. MacKenzie, S. B. (1986). The role of attention in mediating the effect of advertising on attribute importance. Journal o f Consumer Research, 13, 174-195. MacKenzie, S. B., & Spreng, R. A. (1992). How does motivation moderate the impact of central and peripheral processing on brand attitudes and intentions? Journal o f Consumer Research, 18, 519-529. Mackie, D. M. (1987). Systematic and nonsystematic processing of majority and minority persuasive communications. Journal o f Personality and Social Psychology, 53, 41-52. Mackie, D. M ., & Asuncion, A. G. (1990). On-line and memory-based modification of attitudes: Determinants of message recall-attitude change correspondence. Journal o f Personality and Social Psychology, 59, 5-16. Mackie, D. M ., Asuncion, A. G., & Rosselli, F. (1992). The impact of positive affect on persuasion processes. In M. S. Clark (Ed.), Emotion and social behavior (Vol. 14, pp. 247-270). Newbury Park, CA: Sage. Mackie, D. M ., & Worth, L. T. (1989). Processing deficits and the mediation of positive affect in persuasion. Journal o f Personality and Social Psychology, 57, 27-40. Mackie, D. M ., & Worth, L. T. (1990). The impact of distraction on the processing of category-based and attribute-based evaluations. Basic and A pplied Social Psychology, 11, 255-271. Mackie, D. M., & Worth, L. T. (1991). Feeling good, but not thinking straight: The impact of positive mood on persuasion. In J. Forgas (Ed.), Emotion and social judgm ent (pp. 201-220). Oxford, England: Pergamon. Mackie, D. M., Worth, L. T., & Asuncion, A. G. (1990). Processing of persuasive in-group messages. Journal o f Personality and Social Psychology, 58, 812-822.

132

PETTY, PRIESTER, WEGENER

Maheshwaran, D., & Chaiken, S. (1991). Promoting systematic processing in low-motivation set­ tings: Effect of incongruent information on processing and judgment. Journal o f Personality and Social Psychology, 61, 13-33. Mandler, G. (1982). The structure of value: Accounting for taste. In M. S. Clark & S. T. Fiske (Eds.), Affect and cognition: The 17th annual Carnegie symposium (pp. 3-36). Hillsdale, NJ: Lawrence Erlbaum Associates. Markus, H. (1977). Self-schemata and processing information about the self. Journal o f Personality and Social Psychology, 35, 63-78. Martin, L. L. (1986). Set/reset: Use and disuse of concepts in impression formation. Journal o f P er­ sonality and Social Psychology, 51, 493-504. Marwell, G., Aiken, M ., & Demerath, N. J. (1987). The persistence of political attitudes among 1960s civil rights activists. Public Opinion Quarterly, 51, 359-375. Mathur, M ., & Chattopadhyay, A. (1991). The impact of moods generated by television programs on responses to advertising. Psychology and Marketing, 8, 59-77. Matlin, M. W. (1970). Response competition as a mediating factor in the frequency-affect relation­ ship. Journal o f Personality and Social Psychology, 16, 536-552. Mayer, J. D., Gaschke, Y. N., Bravermen, D. L., & Evans, T. W. (1992). Mood-congruent judg­ ment is a general effect. Journal o f Personality and Social Psychology, 63, 119-132. Mazursky, D., & Schul, Y. (1988). The effects of advertisement encoding on the failure to discount information: Implications for the sleeper effect. Journal o f Consumer Research, 15, 24-36. McCann, C. D., & Higgins, E. T. (1984). Individual differences in communication: Social cognitive determinants and consequences. In H. E. Sypher & J. L. Applegate (Eds.), Understanding inter­ personal communication: Social cognitive and strategic processes in children and adults (pp. 172-210). Beverly Hills, CA: Sage. McCann, C. D., Higgins, E. T ., & Fondacaro, R. A. (1991). Primacy and recency in communica­ tion and self-persuasion: How successive audiences and multiple encodings influence subsequent evaluative judgments. Social Cognition, 9, 47-66. McGuire, W. J. (1960a). Cognitive consistency and attitude change. Journal o f Abnormal and Social Psychology, 60, 345-353. McGuire, W. J. (1960b). A syllogistic analysis of cognitive relationships. In C. I Hovland & M. J. Rosenberg (Eds.), Attitude organization and change: An analysis o f consistency among atti­ tude components (pp. 65-111). New Haven, CT: Yale University Press. McGuire, W. J. (1964). Inducing resistance to persuasion: Some contemporary approaches. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 1, pp. 191-229). New York: Academic Press. McGuire, W. J. (1968). Personality and attitude change: An information-processing theory. In A. G. Greenwald, T. C. Brock, & T. M. Ostrom (Eds.), Psychological foundations o f attitudes (pp. 171-196). New York: Academic Press. McGuire, W. J. (1969). The nature of attitudes and attitude change. In G. Lindzey & E. Aronson (Eds.), Handbook o f social psychology (Vol. 3, pp. 136-314). Reading, MA: Addison-Wesley. McGuire, W. J. (1981). The probabilogical model of cognitive structure and attitude change. In R. E. Petty, T. M. Ostrom, & T. C. Brock (Eds.), Cognitive responses in persuasion (pp. 291-307). Hillsdale, NJ: Lawrence Erlbaum Associates. McGuire, W. J. (1985). Attitudes and attitude change. In G. Lindzey & E. Aronson (Eds.), Hand­ book o f social psychology (Vol. 2, pp. 233-346). New York: Random House. McGuire, W. J. (1989). Theoretical foundations of campaigns. In R. E. Rice & C. K. Atkin (Eds.), Public communication campaigns (pp. 43-66). Newbury Park, CA: Sage. McGuire, W. J., & McGuire, C. V. (1991). The content, structure, and operation of thought sys­ tems. In R. S. Wyer, Jr. & T. Srull (Eds.), Advances in social cognition (Vol. 4, pp. 1-78). Hills­ dale, NJ: Lawrence Erlbaum Associates. McGuire, W. J., & McGuire, C. V. (1992). Cognitive versus affective positive asymmetries in thought systems. European Journal o f Social Psychology, 22, 571-591.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

133

McGuire, W. J., & Papageorgis, D. (1961). The relative efficacy of various types of prior beliefdefense in producing immunity against persuasion. Journal o f Abnormal and Social Psychology, 62, 327-337. McSweeney, F. K., & Bierly, C. (1984). Recent developments in classical conditioning. Journal o f Consumer Research, 11, 619-631. Meyers-Levy, J., & Tybout, A. (1989). Schema-congruity as a basis for product evaluation. Journal o f Consumer Research, 16, 39-54. Millar, M. G., & Millar, K. U. (1990). Attitude change as a function of attitude type and argument type. Journal o f Personality and Social Psychology, 59, 217-228. Millar, M. G., & Tesser, A. (1986a). Effects of affective and cognitive focus on the attitude-behavior relation. Journal o f Personality and Social Psychology, 51, 270-276. Millar, M. G., & Tesser, A. (1986b). Thought-induced attitude change: The effects of schema struc­ ture and commitment. Journal o f Personality and Social Psychology, 51, 259-269. Millar, M. G., & Tesser, A. (1992). The role of beliefs and feelings in guiding behavior: The mis­ match model. In C. Martin & A. Tesser (Eds.), Construction o f social judgm ent (pp. 277-300). Hillsdale, NJ: Lawrence Erlbaum Associates. Miller, A. G., McHoskey, J. W ., Bane, C. M., & Dowd, T. G. (1993). The attitude polarization phenomenon: Role of response measure, attitude extremity, and behavioral consequences of reported attitude change. Journal o f Personality and Social Psychology, 64, 561-574. Miniard, P. W ., Bhatla, S., Lord, K. R., Dickson, P. R., & Unnava, H. R. (1991). Picture-based persuasion processes and the moderating role of involvement. Journal o f Consumer Research, 18, 92-107. Miniard, P. W ., Bhatla, S., & Sirdeshmukh, D. (1992). Mood as a determinant of postconsumption product evaluations: Mood effects and their dependency on the affective intensity of the consump­ tion experience. Journal o f Consumer Psychology, 1, 173-195. Miniard, P. W ., Sirdeshmukh, D., & Innis, D. E. (1992). Peripheral persuasion and brand choice. Journal o f Consumer Research, 19, 226-239. Mitnick, L., & McGinnies, E. (1958). Influencing ethnocentrism in small discussion groups through a film communication. Journal o f Abnormal and Social Psychology, 56, 82-92. Moore, D. L., Hausknecht, D., & Thamodaran, K. (1986). Time compression, response opportuni­ ty, and persuasion. Journal o f Consumer Research, 13, 85-99. Moore, D. L., & Reardon, R. (1987). Source magnification: The role of multiple sources in the processing of advertising appeals. Journal o f Marketing Research, 24, 412-417. Moscovici, S. (1980). Toward a theory of conversion behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 13, pp. 209-239). New York: Academic Press. Moscovici, S., Lage, E., & Naffrechoux, M. (1969). Influence of a consistent minority on the responses of a majority in a color perception task. Sociometry, 32, 365-380. Moscovici, S., & Neve, P. (1973). Studies in social influence: II. Instrumental and symbolic influence. European Journal o f Social Psychology, 3, 461-471. Mugny, G., & Perez, J. A. (1991). The social psychology o f minority influence. (V. W. Lamongie, Trans.) New York: Cambridge University Press. Munch, J. M ., & Swasy, J. L. (1988). Rhetorical question, summarization frequency, and argument strength effects on recall. Journal o f Consumer Research, 15, 69-76. Murphy, S. T., & Zajonc, R. B. (1993). Affect, cognition, and awareness: Affective priming with optimal and suboptimal exposures. Journal o f Personality and Social Psychology, 64, 723-739. Murray, N., Sujan, H ., Hirt, E. R., & Sujan, M. (1990). The influence of mood on categorization: A cognitive flexibility interpretation. Journal o f Personality and Social Psychology, 59, 411-425. Norman, R. (1975). Affective-cognitive consistency, attitudes, conformity, and behavior. Journal o f Personality and Social Psychology, 32, 83-91. Obermiller, C. (1985). Varieties of mere exposure: The effects of processing style and repetition on affective response. Journal o f Consumer Research, 12, 17-30. Olson, J., & Zanna, M. (1993). Attitude and attitude change. Annual Review o f Psychology, 44, 117-154.

134

PETTY, PRIESTER, WEGENER

Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement o f meaning. Urbana, IL: University of Illinois Press. Osgood, C. E., & Tannenbaum, P. H. (1955). The principle of congruity in the prediction of attitude change. Psychological Review, 62, 42-55. Osterhouse, R. A ., & Brock, T. C. (1970). Distraction increases yielding to propaganda by inhibit­ ing counterarguing. Journal o f Personality and Social Psychology, 15, 344-358. Pallak, S. R. (1983). Salience of a communicator’s physical attractiveness and persuasion: A heuris­ tic versus systematic processing interpretation. Social Cognition, 2, 158-170. Peak, H. (1955). Attitude and motivation. In M. R. Jones (Ed.), Nebraska symposium on motivation (Vol. 3, pp. 149-188). Lincoln, NE: University of Nebraska Press. Pelham, B. W. (1991). On confidence and consequence: The certainty and importance of selfknowledge. Journal o f Personality and Social Psychology, 60, 518-530. Perloff, R. M ., & Brock, T. C. (1980). And thinking makes it so: Cognitive responses to persua­ sion. In M. Roloff & G. Miller (Eds.), Persuasion: N ew directions in theory and research (pp. 67-100). Beverly Hills, CA: Sage. Petty, R. E. (in press). Two routes to persuasion: State of the art. In G. d’Ydewalle, P. Eelen, & P. Bertelson (Eds.), Current advances in psychological science: An international perspective. Hills­ dale, NJ: Lawrence Erlbaum Associates. Petty, R. E ., & Cacioppo, J. T. (1977). Forewarning, cognitive responding, and resistance to per­ suasion. Journal o f Personality and Social Psychology, 35, 645-655. Petty, R. E., & Cacioppo, J. T. (1979a). Effects of forewarning of persuasive intent and involve­ ment on cognitive responses. Personality and Social Psychology Bulletin, 5, 173-176. Petty, R. E., & Cacioppo, J. T. (1979b). Issue-involvement can increase or decrease persuasion by enhancing message-relevant cognitive responses. Journal o f Personality and Social Psychology, 37, 1915-1926. Petty, R. E., & Cacioppo, J. T. (1981). Attitudes and persuasion: Classic and contemporary ap­ proaches. Dubuque, IA: Wm. C. Brown. Petty, R. E., & Cacioppo, J. T. (1983). The role of bodily responses in attitude measurement and change. In J. T. Cacioppo & R. E. Petty (Eds.), Social psychophysiology: A sourcebook (pp. 51-101). New York: Guilford. Petty, R. E., & Cacioppo, J. T. (1984a). The effects of involvement on responses to argument quan­ tity and quality: Central and peripheral routes to persuasion. Journal o f Personality and Social Psychology, 46, 69-81. Petty, R. E., & Cacioppo, J. T. (1984b). Source factors and the Elaboration Likelihood Model of persuasion. Advances in Consumer Research, 11, 668-672. Petty, R. E., & Cacioppo, J. T. (1986a). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlag. Petty, R. E., & Cacioppo, J. T. (1986b). The Elaboration Likelihood Model of persuasion. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 19, pp. 123-205). New York: Academic Press. Petty, R. E., & Cacioppo, J. T. (1990). Involvement and persuasion: Tradition versus integration. Psychological Bulletin, 107, 367-374. Petty, R. E., Cacioppo, J. T., & Haugtvedt, C. (1992). Involvement and persuasion: An apprecia­ tive look at the Sherifs’ contribution to the study of self-relevance and attitude change. In D. Granberg & G. Sarup (Eds.), Social judgm ent and inter group relations: Essays in honor o f Muzifer Sherif {pp. 147-175). New York: Springer-Verlag. Petty, R. E., Cacioppo, J. T ., & Heesacker, M. (1981). Effects of rhetorical questions on persua­ sion: A cognitive response analysis. Journal o f Personality and Social Psychology, 40, 432-440. Petty, R. E., Cacioppo, J. T ., & Kasmer, J. A. (1988). The role of affect in the Elaboration Likeli­ hood Model of persuasion. In L. Donohew, H. E. Sypher, & E. T. Higgins (Eds.), Communica­ tion, social cognition and affect (pp. 117-146). Hillsdale, NJ: Lawrence Erlbaum Associates.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

135

Petty, R. E., Cacioppo, J. T ., & Schumann, D. W. (1983). Central and peripheral routes to advertis­ ing effectiveness: The moderating role of involvement. Journal o f Consumer Research, 10, 135-146. Petty, R. E., Gleicher, F., & Baker, S. M. (1991). Multiple roles for affect in persuasion. In J. Forgas (Ed.), Emotion and social judgm ents (pp. 181-200). Oxford, England: Pergamon. Petty, R. E ., Harkins, S. G., & Williams, K. D. (1980). The effects of group diffusion of cognitive effort on attitudes: An information processing view. Journal o f Personality and Social Psycholo­ gy, 38, 81-92. Petty, R. E., Haugtvedt, C. P., & Smith, S. M. (in press). Elaboration as a determinant of attitude strength. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Petty, R. E., Kasmer, J. A., Haugtvedt, C. P., & Cacioppo, J. T. (1987). A reply to Stiff and Boster. Communication Monographs, 54, 257-263. Petty, R. E., & Krosnick, J. A. (Eds.) (in press). Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Petty, R. E., Ostrom, T. M ., & Brock, T. C. (Eds.) (1981). Cognitive responses in persuasion. Hillsdale, NJ: Lawrence Erlbaum Associates. Petty, R. E., Schumann, D. W ., Richman, S. A., & Strathman, A. J. (1993). Positive mood and persuasion: Different roles for affect under high- and low-elaboration conditions. Journal o f P er­ sonality and Social Psychology, 64, 5-20. Petty, R. E., & Wegener, D. T. (1991). Thought systems, argument quality, and persuasion. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Advances in social cognition (Vol. 4, pp. 147-161). Hills­ dale, NJ: Lawrence Erlbaum Associates. Petty, R. E., & Wegener, D. T. (1993). Flexible correction processes in social judgment: Correct­ ing for context induced contrast. Journal o f Experimental Social Psychology, 29, 137-165. Petty, R. E., Wegener, D. T ., Fabrigar, L. R., Priester, J. R., & Cacioppo, J. T. (in press). Con­ ceptual and methodological issues in the elaboration likelihood model of persuasion: A reply to the Michigan State critics. Communication Theory. Petty, R. E ., Wells, G. L., & Brock, T. C. (1976). Distraction can enhance or reduce yielding to propaganda: Thought disruption versus effort justification. Journal o f Personality and Social P sy­ chology, 34, 874-884. Pfau, M., & Burgoon, M. (1988). Inoculation in political campaign communication. Human Com­ munication Research, 15, 91-111. Pfau, M., Kenski, H. C., Nitz, M., & Sorenson, J. (1990). Efficacy of inoculation strategies in promot­ ing resistance to political attack messages: Application to direct mail. Communication Monographs, 57, 25-43. Pieters, R. G. M. (1988). Attitude-behavior relationships. In W. F. V. Raaij, G. M. V. Veldhoven, & K. Warneryd (Eds.), Handbook o f economic psychology (pp. 147-204). Dordrecht: Kluwer Academic. Pittman, T. S. (in press). Control motivation and attitude change. In G. Weary, F. Gleicher, & K. Marsh (Eds.), Control motivation and social cognition. New York: Springer-Verlag. Pratkanis, A. R. (1989). The cognitive representation of attitudes. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 71-98). Hillsdale, NJ: Lawrence Erlbaum Associates. Pratkanis, A. R., Breckler, S. J., & Greenwald, A. R. (Eds.). (1989). Attitude structure and func­ tion. Hillsdale, NJ: Lawrence Erlbaum Associates. Pratkanis, A. R., & Greenwald, A. G. (1989). A sociocognitive model of attitude structure and func­ tion. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22, pp. 245-286). San Diego, CA: Academic Press. Pratkanis, A. R., Greenwald, A. G., Leippe, M. R., & Baumgardner, M. H. (1988). In search of reliable persuasion effects: III. The sleeper effect is dead. Long live the sleeper effect. Journal o f Personality and Social Psychology, 54, 203-218.

136

PETTY, PRIESTER, WEGENER

Prentice, D. A. (1987). Psychological correspondence of possessions, attitudes, and values. Journal o f Personality and Social P sychology, 53, 993-1003. Priester, J. R., & Petty, R. E. (1991, May). The role o f need fo r cognition in the attributional analy­ sis o f persuasion. Paper presented at the meeting of the Midwestern Psychological Association, Chicago. Priester, J. R., & Petty, R. E. (1993). Perceived source trustworthiness as a determinant o f message scrutiny. Unpublished manuscript, Ohio State University, Columbus. Puckett, J. M ., Petty, R. E., Cacioppo, J. T ., & Fisher, D. L. (1983). The relative impact of age and attractiveness stereotypes on persuasion. Journal o f Gerontology, 38, 340-343. Raden, D. (1985). Strength-related attitude dimensions. Social Psychology Quarterly, 48, 312-330. Ratneshwar, S., & Chaiken, S. (1991). Comprehension’s role in persuasion: The case of its mod­ erating effect on the persuasive impact of source cues. Journal o f Consumer Psychology, 18, 52-62. Razran, G. H. S. (1940). Conditioned response changes in rating and appraising sociopolitical slo­ gans. Psychological Bulletin, 37, 481. Reardon, R., & Rosen, S. (1984). Psychological differentiation and the evaluation of juridic infor­ mation: Cognitive and affective consequences. Journal o f Research in Personality, 18, 195-211. Regan, D. T ., & Fazio, R. H. (1977). On the consistency between attitudes and behavior: Look to the method of attitude formation. Journal o f Experimental Social Psychology, 13, 28-45. Rennier, G. A. (1988). The strength o f the object-evaluation association, the attitude-behavior rela­ tionship and the elaboration likelihood m odel o f persuasion. Unpublished doctoral dissertation, University of Missouri, Columbia. Rethans, A. J., Swasy, J. L., & Marks, L. J. (1986). Effects of television commercial repetition, receiver knowledge, and commercial length: A test of the two-factor model. Journal o f Marketing Research, 23, 50-61. Rhodes, N., & Wood, W. (1992). Self-esteem and intelligence affect influenceability: The mediat­ ing role of message reception. Psychological Bulletin, 111, 156-171. Roberts, J. V. (1985). The attitude-memory relationship after 40 years: A meta-analysis of the liter­ ature. Basic and Applied Social Psychology, 6, 221-241. Rogers, R. W. (1983). Cognitive and physiological processes in fear appeals and attitude change: A revised theory of protection motivation. In J. T. Cacioppo & R. E. Petty (Eds.), Social p sy ­ chophysiology: A sourcebook (pp. 153-176). New York: Guilford. Rokeach, M. (1960). The open and closed mind. New York: Basic Books. Rokeach, M. (1968). Beliefs, attitudes, and values: A theory o f organization and change. San Fran­ cisco: Jossey-Bass. Ronis, D. L., Baumgardner, M. H., Leippe, M. R., Cacioppo, J. T ., & Greenwald, A. G. (1977). In search of reliable persuasion effects: I. A computer-controlled procedure for studying persua­ sion. Journal o f Personality and Social Psychology, 35, 548-569. Ronis, D. L., Yates, J. F., & Kirscht, J. P. (1989). Attitudes, decisions, and habits as determinants of repeated behavior. In A. Pratkanis, S. Breckler, & A. Greenwald (Eds.), Attitude structure and function (pp. 213-239). Hillsdale, NJ: Lawrence Erlbaum Associates. Rosenberg, M. J. (1956). Cognitive structure and attitudinal affect. Journal o f Abnormal and Social Psychology, 53, 367-372. Rosenberg, M. J. (1960). An analysis of affective-cognitive consistency. In C. I. Hovland & M. J. Rosenberg (Eds.), Attitude organization and change: An analysis o f consistency among atti­ tude components (pp. 15-64). New Haven, CT: Yale University Press. Rosenberg, M. J., & Hovland, C. I. (1960). Cognitive, affective, and behavioral components of atti­ tudes. In C. I. Hovland & M. J. Rosenberg (Eds.), Attitude organization and change: An analysis o f consistency among attitude components (pp. 1-14). New Haven, CT: Yale University Press. Roskos-Ewoldsen, D. R., & Fazio, R. H. (1992a). The accessibility of source likability as a deter­ minant of persuasion. Personality and Social Psychology Bulletin, 18, 19-25.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

137

Roskos-Ewoldsen, D. R., & Fazio, R. H. (1992b). On the orienting value of attitudes: Attitude ac­ cessibility as a determinant of an object’s attraction of visual attention. Journal o f Personality and Social Psychology, 63, 198-211. Ross, M ., McFarland, C ., Conway, M ., & Zanna, M. P. (1983). Reciprocal relation between atti­ tudes and behavior recall: Committing people to newly formed attitudes. Journal o f Personality and Social Psychology, 45, 257-267. Sadler, O., & Tesser, A. (1973). Some effects of salience and time upon interpersonal hostility and attraction. Sociometry, 36, 99-112. Sanbonmatsu, D. M ., & Fazio, R. H. (1990). The role of attitudes in memory-based decision mak­ ing. Journal o f Personality and Social Psychology, 59, 614-622. Sanbonmatsu, D. M ., & Kardes, F. R. (1988). The effects of physiological arousal on information processing and persuasion. Journal o f Consumer Research, 15, 379-385. Sawyer, A. G. (1981). Repetition, cognitive responses, and persuasion. In R. E. Petty, T. M. Os­ trom, & T. C. Brock (Eds.), Cognitive responses in persuasion (pp. 237-261). Hillsdale, NJ: Lawrence Erlbaum Associates. Sawyers, B. K., & Anderson, N. H. (1971). Test of integration theory in attitude change. Journal o f Personality and Social Psychology, 18, 230-233. Scher, S. J., & Cooper, J. (1989). Motivational basis of dissonance: The singular role of behavioral consequences. Journal o f Personality and Social Psychology, 56, 899-906. Schmidt, D. F., & Sherman, R. C. (1984). Memory for persuasive messages: A test of a schemacopy-plus-tag model. Journal o f Personality and Social Psychology, 47, 17-25. Schul, Y., & Burnstein, E. (1983). The informational basis of social judgments: Memory for infor­ mative and uninformative arguments. Journal o f Experimental Social Psychology, 19, 422-433. Schul, Y., & Burnstein, E. (1985). When discounting fails: Conditions under which individuals use discredited information in making a judgment. Journal o f Personality and Social Psychology, 49, 894-903. Schumann, D. W ., Petty, R. E., & Clemons, D. S. (1990). Predicting the effectiveness of different strategies of advertising variation: A test of the repetition-variation hypotheses. Journal o f Con­ sumer Research, 17, 192-202. Schumann, D. W ., & Thorson, E. (1990). The influence of viewing context on commercial effec­ tiveness. Current Issues and Research in Advertising, 12, 1-24. Schwarz, N. (1990). Feelings as information: Informational and motivational functions of affective states. In E. T. Higgins & R. M. Sorrentino (Eds.), Handbook o f motivation and cognition: Foun­ dations o f social behavior (Vol. 2, pp. 527-561). New York: Guilford. Schwarz, N., Bless, H., & Bohner, G. (1991). Mood and persuasion: Affective states influence the processing of persuasive communications. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 161-201). San Diego: Academic Press. Schwarz, N., Servay, W ., & Kumpf, M. (1985). Attribution of arousal as a mediator of the effec­ tiveness of fear-arousing communications. Journal o f Applied Social Psychology, 15, 74-84. Scott, C. A., & Yalch, R. J. (1978). A test of the self-perception explanation of the effects of re­ wards on intrinsic interest. Journal o f Experimental Social Psychology, 14, 180-192. Scott, W. A. (1968). Attitude measurement. In G. Lindzey & E. Aronson (Eds.), Handbook o f so­ cial psychology (Vol. 2, pp. 204-273). Reading, MA: Addison-Wesley. Shavitt, S. (1989). Operationalizing functional theories of attitude. In A. R. Pratkanis, S. J. Breck­ ler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 311-338). Hillsdale, NJ: Lawrence Erlbaum Associates. Shavitt, S. (1990). The role of attitude objects in attitude functions. Journal o f Experimental Social Psychology, 26, 124-148. Shavitt, S., & Brock, T. C. (1986). Self-relevant responses in commercial persuasion: Field and ex­ perimental tests. In J. Olson & K. Sentis (Eds.), Advertising and consumer psychology (Vol. 3, pp. 149-171). New York: Praeger.

138

PETTY, PRIESTER, WEGENER

Shavitt, S., & Fazio, R. H. (1991). Effects of attribute salience on the consistency between attitudes and behavior predictions. Personality and Social Psychology Bulletin, 17, 507-516. Sherer, M., & Rogers, R. W. (1984). The role of vivid information in fear appeals and attitude change. Journal o f Research in Personality, 18, 321-334. Sherif, C. W ., Sherif, M., & Nebergall, R. E. (1965). Attitude and attitude change: The social judgmentinvolvement approach. Philadelphia: Saunders. Sherif, M ., & Hovland, C. I. (1961). Social judgment: Assimilation and contrast effects in communi­ cation and attitude change. New Haven, CT: Yale University Press. Sherif, M ., & Sherif, C. W. (1967). Attitude as the individual’s own categories: The social judgmentinvolvement approach to attitude and attitude change. In C. W. Sherif & M. Sherif (Eds.), Atti­ tude, ego-involvement and change (pp. 105-139). New York: Wiley. Sherman^ S. J., Cialdini, R. B., Schwartzman, D. F., & Reynolds, K. D. (1985). Imagining can heighten or lower the perceived likelihood of contracting a disease: The mediating effect of ease of imagery. Personality and Social Psychology Bulletin, 11, 118-127. Sherman, S. J., Mackie, D. M ., & Driscoll, D. M. (1990). Priming and the differential use of dimen­ sions in evaluation. Personality and Social Psychology Bulletin, 16, 405-418. Sherman, S. J., Zehner, K. S., Johnson, J., & Hirt, E. R. (1983). Social explanation: The role of timing, set, and recall on subjective likelihood estimates. Journal o f Personality and Social P sy­ chology, 44, 1127-1143. Shimp, T. A., Stuart, E. W ., & Engle, R. W. (1991). A program of classical conditioning experi­ ments testing variations in the conditioned stimulus and context. Journal o f Consumer Research, 18, 1-12. Sinclair, R. C., & Mark, M. M. (1992). The influence of mood state on judgment and action: Ef­ fects on persuasion, categorization, social justice, person perception, and judgmental accuracy. In L. L. Martin & A. Tesser (Eds.), The construction o f social judgm ents (pp. 165-193). Hills­ dale, NJ: Lawrence Erlbaum Associates. Sivacek, J., & Crano, W. D. (1982). Vested interest as a moderator of attitude-behavior consisten­ cy. Journal o f Personality and Social Psychology, 43, 210-221. Slamecka, N. J., & Graf, P. (1978). The generation effect: Delineation of a phenomenon. Journal o f Experimental Psychology: Human Learning and Memory, 4, 592-604. Smith, M. B., Bruner, J. S., & White, R. W. (1956). Opinions and personality. New York: Wiley. Smith, S. M ., & Shaffer, D. R. (1991). Celerity and cajolery: Rapid speech may promote or inhibit persuasion through its impact on message elaboration. Personality and Social Psychology Bulle­ tin, 17, 663-669. Snyder, M. (1979). Self-monitoring processes. In L. Berkowitz (Ed.), Advances in experimental so­ cial psychology (Vol. 12, pp. 86-128). New York: Academic Press. Snyder, M ., & DeBono, K. G. (1985). Appeals to images and claims about quality: Understanding the psychology of advertising. Journal o f Personality and Social Psychology, 49, 586-597. Snyder, M ., & DeBono, K. G. (1987). A functional approach to attitudes and persuasion. In M. P. Zanna, J. M. Olson, & C. P. Herman (Eds.), Social influence: The Ontario Symposium (Vol. 5, pp. 107-125). Hillsdale, NJ: Lawrence Erlbaum Associates. Snyder, M ., & DeBono, K. G. (1989). Understanding the functions of attitudes: Lessons from per­ sonality and social behavior. In S. J. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Atti­ tude structure and function (pp. 339-359). Hillsdale, NJ: Lawrence Erlbaum Associates. Snyder, M., & Swann, W. B. (1976). When actions reflect attitudes: The politics of impression manage­ ment. Journal o f Personality and Social Psychology, 34, 1034-1042. Srull, T. K. (1983). The role of prior knowledge in the acquisition, retention, and use of new infor­ mation. Advances in Consumer Research, 10, 572-576. Staats, A. W ., & Staats, C. K. (1958). Attitudes established by classical conditioning. Journal o f Abnormal and Social Psychology, 57, 37-40. Staats, A. W ., Staats, C. K., & Crawford, H. L. (1962). First-order conditioning of meaning and the parallel conditioning of a GSR. Journal o f General Psychology, 67, 159-167.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

139

Stangor, C., Sullivan, L. A., & Ford, T. E. (1991). Affective and cognitive determinants of prejudice. Social Cognition, 9, 359-380. Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261-302). New York: Academic Press. Steele, C. M ., Spencer, S. J., & Lynch, M. (1993). Self-image resilience and dissonance: The role of affirmational resources. Journal o f Personality and Social Psychology, 6 , 885-896. Stiff, J. B. (1986). Cognitive processing of persuasive message cues: A meta-analytic review of the effects of supporting information on attitudes. Communication Monographs, 53, 75-89. Strack, F., Martin, L., & Stepper, S. (1988). Inhibiting and facilitating conditions of the human smile: A nonobtrusive test of the facial feedback hypothesis. Journal o f Personality and Social Psycholo­ gy, 54, 768-777. Strack, F., Schwarz, N ., Bless, H., Kiibler, A., & Wanke, M. (1993). Awareness of the influence as a determinant of assimilation versus contrast. European Journal o f Social Psychology, 23, 53-62. Stroebe, W ., & Diehl, M. (1988). When social support fails: Supporter characteristics in complianceinduced attitude change. Personality and Social Psychology Bulletin, 14, 136-144. Strong, E. K. (1925). The psychology o f selling and advertising. New York: McGraw-Hill. Stuart, E. W ., Shimp, T. A., & Engle, R. W. (1987). Classical conditioning of consumer attitudes: Four experiments in an advertising context. Journal o f Consumer Research, 14, 334-349. Swann, W. B., Hixon, J. G., Stein-Seroussi, A., & Gilbert, D. T. (1990). The fleeting gleam of praise: Cognitive processes underlying behavioral reactions to self-relevant feedback. Journal o f Personality and Social Psychology, 59, 17-26. Swasy, J. L., & Munch, J. M. (1985). Examining the target of receiver elaborations: Rhetorical question effects on source processing and persuasion. Journal o f Consumer Research, 11, 877-886. Tanford, S., & Penrod, S. (1984). Social influence model: A formal integration of research on majority and minority influence processes. Psychological Bulletin, 95, 189-225. Taylor, S. E. (1975). On inferring one’s attitude from one’s behavior: Some delimiting conditions. Journal o f Personality and Social Psychology, 31, 126-131. Taylor, S. E. (1981). The interface of cognitive and social psychology. In J. H. Harvey (Ed.), Cog­ nition, social behavior, and the environment (pp. 189-211). Hillsdale, NJ: Lawrence Erlbaum Associates. Tesser, A. (1978). Self-generated attitude change. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 11, pp. 289-338). New York: Academic Press. Tesser, A. (1993). The importance of heritability in psychological research: The case of attitudes. Psychological Review, 100, 129-142. Tesser, A., & Cornell, D. P. (1991). On the confluence of self-processes. Journal o f Experimental Social Psychology, 27, 501-526. Tesser, A., & Cowan, C. (1975). Some effects of thought and number of cognitions on attitude change. Social Behavior and Personality, 3, 165-173. Tesser, A., & Leone, C. (1977). Cognitive schemas and thought as determinants of attitude change. Journal o f Experimental Social Psychology, 13, 340-356. Tesser, A., Martin, L ., & Mendolia, M. (in press). The impact of thought on attitude extremity and attitude-behavior consistency. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antece­ dents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Tesser, A., & Shaffer, D. R. (1990). Attitudes and attitude change. Annual Review o f Psychology, 41, 479-523. Tetlock, P. E. (1983). Accountability and complexity of thought. Journal o f Personality and Social Psychology, 45, 74-83. Tetlock, P. E., Skitka, L., & Boettger, R. (1989). Social and cognitive strategies for coping with accountability: Conformity, complexity, and bolstering. Journal o f Personality and Social P sy­ chology, 57, 632-640.

140

PETTY, PRIESTER, WEGENER

Thibodeau, R., & Aronson, E. (1992). Taking a closer look: Reasserting the role of the self-concept in dissonance theory. Personality and Social Psychology Bulletin, 18, 591-602. Thompson, M. M ., Zanna, M. P., & Griffin, D. W. (in press). Let’s not be indifferent about (attitudinal) ambivalence. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Thomsen, C. J., Borgida, E ., & Lavine, H. (in press). The causes and consequences of personal involvement. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and conse­ quences. Hillsdale, NJ: Lawrence Erlbaum Associates. Tice, D. M. (1992). Self-concept change and self-presentation: The looking glass self is also a mag­ nifying glass. Journal o f Personality and Social Psychology, 63, 435-451. Titchener, E. B. (1910). Textbook o f psychology. New York: Macmillan. Tom, G., Pettersen, P., Lau, T ., Burton, T ., & Cook, J. (1991). The role of overt head movement in the formation of affect. Basic and A pplied Social Psychology, 12, 281-289. Tourangeau, R., & Rasinski, K. A. (1988). Cognitive processes underlying context effects in atti­ tude measurement. Psychological Bulletin, 103, 299-314. Tourangeau, R., Rasinski, K. A., & D’Andrade, R. (1991). Attitude structure and belief accessibili­ ty. Journal o f Experimental Social Psychology, 27, 48-75. Triandis, H. C. (1977). Interpersonal behavior. Monterey, CA: Brooks/Cole. Trost, M. R., Maass, A., & Kenrick, D. T. (1992). Minority influence: Personal relevance biases cognitive processes and reverses private acceptance. Journal o f Experimental Social Psychology, 28, 234-254. Tybout, A. M ., & Scott, C. A. (1983). Availability of well-defined internal knowledge and the atti­ tude formation process: Information aggregation versus self-perception. Journal o f Personality and Social Psychology, 44, 474-491. Verplanken, B. (1991). Persuasive communication of risk information: A test of cue versus message processing effects in a field experiment. Personality and Social Psychology Bulletin, 17, 188-193. Watts, W. A. (1967). Relative persistence of opinion change induced by active compared to passive participation. Journal o f Personality and Social Psychology, 5, 4-15. Watts, W. A., & Holt, L. E. (1979). Persistence of opinion change induced under conditions of fore­ warning and distraction. Journal o f Personality and Social Psychology, 37, 778-789. Weber, S. J. (1972). Opinion change is a function o f the associative learning o f content and source factors. Unpublished doctoral dissertation, Northwestern University, Evanston, IL. Wegener, D. T ., Petty, R. E., & Klein, D. J. (in press). Effects of mood on high elaboration attitude change: The mediating role of likelihood judgments. European Journal o f Social Psychology. Wegner, D. M ., Wenzlaff, R., Kerker, R. M ., & Beattie, A. E. (1981). Incrimination through in­ nuendo: Can media questions become public answers? Journal o f Personality and Social Psychol­ ogy, 40, 822-832. Weldon, D. E., & Malpass, R. S. (1981). Effects of attitudinal, cognitive, and situational variables on recall of biased communications. Journal o f Personality and Social Psychology, 40, 39-52. Wells, G. L ., & Petty, R. E. (1980). The effects of overt head movement on persuasion: Compati­ bility and incompatibility of responses. Basic and A pplied Social Psychology, 1, 219-230. Wilder, D. A. (1990). Some determinants of the persuasive power of in-groups and out-groups: Or­ ganization of information and attribution of independence. Journal o f Personality and Social P sy­ chology, 59, 1202-1213. Wilson, T. D ., Dunn, D. S., Kraft, D., & Lisle, D. J. (1989). Introspection, attitude change, and attitude-behavior consistency: The disruptive effects of explaining why we feel the way we do. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22, pp. 287-343). San Diego: Academic Press. Wilson, T. D ., & Hodges, S. D. (1992). Attitudes as temporary constructions. In A. Tesser & L. Martin (Eds.), The construction o f social judgm ent (pp. 37-65). Hillsdale, NJ: Lawrence Erlbaum Associates.

2.

COGNITIVE PROCESSES IN ATTITUDE CHANGE

141

Wilson, T. D., Kraft, D ., & Dunn, D. S. (1989). The disruptive effects of explaining attitudes: The moderating effect of knowledge about the attitude object. Journal o f Experimental Social Psychol­ ogy, 25, 379-400. Wilson, W. R. (1979). Feeling more than we can know: Exposure effects without learning. Journal o f Personality and Social Psychology, 37, 811-821. Wood, W. (1982). Retrieval of attitude-relevant information from memory: Effects on susceptibility to persuasion and on intrinsic motivation. Journal o f Personality and Social Psychology, 42, 798-910. Wood, W ., & Kallgren, C. A. (1988). Communicator attributes and persuasion: Recipients access to attitude-relevant information in memory. Personality and Social Psychology Bulletin, 14, 172-182. Wood, W ., Kallgren, C. A., & Preisler, R. M. (1985). Access to attitude-relevant information in memory as a determinant of persuasion: The role of message attributes. Journal o f Experimental Social Psychology, 21, 73-85. Wood, W ., Rhodes, N., & Biek, M. (in press). Working knowledge and attitude strength: An infor­ mation processing analysis. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Worth, L. T ., & Mackie, D. M. (1987). Cognitive mediation of positive affect in persuasion. Social Cognition, 5, 76-94. Wright, P. L. (1974). Analyzing media effects on advertising responses. Public Opinion Quarterly, 38, 192-205. Wright, P. L. (1980). Message-evoked thoughts: Persuasion research using thought verbalizations. Journal o f Consumer Research, 7, 151-175. Wu, C., & Shaffer, D. R. (1987). Susceptibility to persuasive appeals as a function of source credi­ bility and prior experience with the attitude object. Journal o f Personality and Social Psychology, 52, 677-688. Wyer, R. S., Jr. (1970). Quantitative prediction of belief and opinion change: A further test of a subjective probability model. Journal o f Personality and Social Psychology, 165, 559-570. Wyer, R. S., Jr. (1973). Category ratings as “subjective expected values”: Implications for attitude formation and change. Psychological Review, 80, 446-467. Wyer, R. S., Jr. (1991). The construction and use of thought systems: Some theoretical ambiguities. In R. S. Wyer & T. K. Srull (Eds.), Advances in social cognition (Vol. 4, pp. 203-214). Hills­ dale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr. (1974). Cognitive organization and change: An information-processing approach. Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., Carlston, D. E., & Hartwick, J. (1979). The role of syllogistic reasoning in infer­ ences based upon new and old information. In R. S. Wyer, Jr. & D. E. Carlston (Eds.), Social cognition, inference, and attribution (pp. 221-274). Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., & Goldberg, L. (1970). A probabilistic analysis of the relationships among beliefs and attitudes. Psychological Review, 77, 100-120. Wyer, R. S., Jr., & Hartwick, J. (1980). The role of information retrieval and conditional inference processes in belief formation and change. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 13, pp. 241-284). San Diego: Academic Press. Wyer, R. S., Jr., & Srull, T. K. (1986). Human cognition in its social context. Psychological Review, 93, 322-359. Wyer, R. S., Jr., & Srull, T. K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., & Srull, T. K. (Eds.) (1991). Advances in social cognition (Vol. 4). Hillsdale, NJ: Lawrence Erlbaum Associates. Yalch, R. F ., & Elmore-Yalch, R. (1984). The effect of numbers on the route to persuasion. Journal o f Consumer Research, 11, 522-527. Zajonc, R. B. (1968a). Attitudinal effects of mere exposure. Journal o f Personality and Social Psy­ chology Monograph Supplements, 9, 1-27.

142

PETTY, PRIESTER, WEGENER

Zajonc, R. B. (1968b). Cognitive theories in social psychology. In G. Lindzey & E. Aronson (Eds.), Handbook o f social psychology (2nd ed., Vol. 1, pp. 320-411). Reading, MA: Addison-Wesley. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35, 151-175. Zajonc, R. B., & Markus, H. (1982). Affective and cognitive factors in preferences. Journal o f Con­ sumer Research, 9, 123-131. Zajonc, R. B., & Markus, H. (1984). Affect and cognition: The hard interface. In C. E. Izard, J. Kagan, & R. B. Zajonc (Eds.), Emotions, cognitions and behavior (pp. 73-102). New York: Cam­ bridge University Press. Zaller, J. R. (1992). The nature and origins o f mass opinion. New York: Cambridge University Press. Zanna, M. P., & Cooper, J. (1974). Dissonance and the pill: An attribution approach to studying the arousal properties of dissonance. Journal o f Personality and Social Psychology, 29, 703-709. Zanna, M. P., Kiesler, C. A., & Pilkonis, P. A. (1970). Positive and negative attitudinal affect es­ tablished by classical conditioning. Journal o f Personality and Social Psychology, 14, 321-328. Zanna, M. P., & Rempel, J. K. (1988). Attitudes: A new look at an old concept. In D. Bar-Tal & A. W. Kruglanski (Eds.), The social psychology o f knowledge (pp. 315-334). Cambridge, Eng­ land: Cambridge University Press.

3

Cognitive Perspective in Political Psychology

Shanto Iyengar University o f California, Los Angeles

Victor Ottati State University of N ew York, Stony Brook

Contents Introduction 143 Three Eras of Research in Political Psychology 144 Personality and Politics 144 Attitudes and Voting 145 Cognition and Information Processing 146 Attribution-Based Models of Voting 147 Assessing Economic Outcomes 148 Attributions of Responsibility 149 Uncertainty and Economic Voting 151 Salience Effects in Public Opinion 154 Television News and Issue Salience 155 Television News and Differential Weighting of Issues in Candidate Evaluation 157 Models of Information Processing 158 Information Processing en Route to Forming Issue Opinions 161 Information Processing Models of Candidate Evaluation 165 Affect and Political Judgment 174 Future Directions 177 References 179

INTRODUCTION The social sciences cross paths all too infrequently. Cross-fertilization between the various disciplines usually occurs only at the level o f basic theoretical frame­ works or methodologies. Superficially, the interdisciplinary field of political psy­ chology may appear to be yet another example o f weak communication (between

143

144

IYENGAR AND OTTATI

the field of psychology as the source discipline and political science as the recep­ tor discipline). That is, the practitioners of political psychology are primarily po­ litical scientists who use the theories and methods of psychology to study phenomena of political relevance. A closer examination of current research efforts in political psychology re­ veals that the field cannot be characterized so easily in terms of disciplinary dominance. Instead, each parent discipline has influenced research in the other. Although political scientists, especially those working in the fields of public opinion and voting, may have been the more eager participants in this joint venture (in­ deed, most graduate training programs in political psychology are housed in depart­ ments of political science), the range of phenomena and activities studied by political scientists and the nonexperimental methods they use increasingly have migrated to psychology. In this sense, political psychology is a dialogue marked by genuine intellectual exchange (Iyengar & McGuire, 1992).

Three Eras of Research in Political Psychology

As McGuire (1992) pointed out in his comprehensive survey of the history of political psychology, the field’s intellectual antecedents encompass three distinct theoretical paradigms of social psychology: personality, attitudes, and informa­ tion processing. Each paradigm spawned a phase of research in political psychol­ ogy that corresponded closely in time to the rise and decline of the underlying theoretical orientation. The earliest phase of research was derived from theories of personality. These studies, which flourished during the 1940s and 1950s, took inspiration from the work of clinicians. The second phase of research, which peaked in the 1960s and 1970s, was grounded in theories of attitude formation and change, and sought to delineate the impact of attitudes on voting. The most recent phase of research (and the subject of this chapter) borrows heavily from work in cognitive-social psychology, and seeks to explain public opinion and po­ litical behavior in terms of information processing.

Personality and Politics

The initial phase of research into the well springs of political behavior was de­ rived primarily from the clinical tradition in psychology, which tended to view behavior as the product of stable personality needs and traits. For political psy­ chologists, the questions of central interest concerned political leadership. Why did people seek political office? What personality traits distinguished political leaders from the citizenry at large? What personality traits created voter appeal? In general, what were the factors or experiences that shaped the political person­ ality? The preferred explanations for these questions were all derived from psy­ choanalytic research. For example, according to Lasswell’s (1948) general theory,

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

145

unresolved anxieties about ego and self-esteem drove individuals to seek public office as a means of coping with their psychic needs. Two schools of thought emerged concerning questions of political personali­ ty, one qualitative and historical, the other more quantitative in method. The former (psycho history) consisted primarily of clinically informed biographies of partic­ ular political figures. For example, in their study of Woodrow Wilson, George and George (1964) traced the president’s rigidity and intransigence during the debate in the U.S. Senate concerning American participation in the League of Nations to his father’s punitive child-rearing practices. Similarly, Barber (1985) proposed the general hypothesis that presidents with relatively low self-esteem are unable to reverse their course while in office. For instance, Lyndon Johnson continued to escalate U.S. military involvement in South Vietnam, despite mount­ ing elite and public opposition to this policy. The quantitative approach to political psychology attempted to measure (typi­ cally with the use of survey methods) the correlates of particular personality traits. Personality generally was specified as the product of interpersonal, intergroup, and general cultural processes. Important examples of this genre included the Adorno, Frenkel-Brunswik, Levinson, & Sanford (1950) The Authoritarian Per­ sonality', which suggested that parental suppression of aggressive impulses produced a buildup of hostility, which was released on out-groups such as ethnic or racial minorities; and McClelland’s (1961) The Achieving Society, which sug­ gested that the development of entrepreneurial roles and skills required a strong need for achievement, and which was the product of cultural and religious values. Attitudes and Voting

The early interest in personality, culture, and socialization led naturally to research on values and attitudes. The guiding theme underlying this phase of political psy­ chology was that people seek to achieve particular psychological outcomes (in­ cluding attitudinal states) that are ego-gratifying, socially adaptive, or otherwise fulfilling. For instance, aspiring doctors may express opinions critical of nation­ alized health insurance schemes and other efforts to regulate the medical profes­ sion. In addition to this functionalist perspective, other arguments were also prevalent. The beliefs-attitudes-actions sequence hypothesized that attitudes ex­ plain individual behavior, and various forms of consistency-balance theory sug­ gested that people consciously or unconsciously formulate a schema of opinions that can be viewed as coherent and systematic, rather than as psychological loose ends. World events during World War II (particularly the phenomenon of ordinary citizens in Europe lending their support to extremist movements) forced attitude researchers to expand their repertoire of explanatory variables. In addition to per­ sonality and cultural factors, the mass media and propaganda campaigns loomed large as contextual sources of political attitudes. With the emergence of television

146

IYENGAR AND OTTATI

in the postwar era, the impact of information on voting and public opinion and the effects of particular campaign messages became important areas of research. The classic studies of Lazarsfeld and his colleagues into the determinants of voting preference illustrated the various cross-currents running through this phase of political psychology (Berelson, Lazarsfeld, & McPhee, 1954; Lazarsfeld, Berelson, & Gaudet, 1949). In keeping with the personality approach, political be­ havior was viewed as dispositional. Group affiliations, ideology, and long-standing political values were seen as crucial to understanding voting. Thus, voters’ party identification, known to be acquired relatively early in life, was considered the paramount electoral force. The expected strain toward attitudinal consistency was thought to dictate that individuals would express attitudes and opinions that were consistent with their party affiliation. Short-term forces, such as information flows, were relevant, but were thought to serve merely to reinforce or activate natural (i.e., partisan) preferences. In sum, researchers typically traced political attitudes to dispositional factors, including cultural values, personality traits, group affiliations, and other rela­ tively stable properties of the individual. What these various arguments had in common was a view of human beings operating under a consistency constraint. Cognition and Information Processing

As discussed earlier, research on all aspects of attitude formation, development, and change preoccupied social psychologists in the period following World War II. In the late 1950s, consistency theories of attitudes, such as dissonance theory, were enriched by the insights of cognitive psychologists. With the ever-increasing influence of attribution theory in social psychology, by the late 1970s, political psychologists began to turn from motivational arguments such as balance and dis­ sonance theory to the insights of cognitive psychology. Much of cognitive psychology suggests mechanistic accounts of political be­ havior that are rooted in the components and structure of the cognitive system. Under the influence of cognitive psychology, the focus of political psychology shifted from an interest in motive-based or dynamic to capacity-based or direc­ tive explanations. The mechanical aspects of human information processing abil­ ity (e.g., the form in which concepts are stored in long-term memory and the processes by which information can be retrieved) emerged as explanatory rivals to the standard dispositional-motivational theories of political behavior. This chapter integrates and synthesizes the research in this most recent phase of political psychology. Rather than undertake an encyclopaedic and chronologi­ cal compilation of the available literature, the focus here is on three principal psychological approaches that command a wide following in political science. First, we describe an extensive body of work based on theories similar to attri­ bution theory. This work argued that voting behavior is largely the product of political (but especially economic) outcomes. This research generally was

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

147

“apsychological,” but political psychologists have begun to specify the processes by which citizens’ assessments of the state of the nation affect their voting preferences. In contemporary democratic societies, information about public affairs is trans­ mitted primarily by television; to examine information processing in the political context necessitates attention to this particular medium of mass communication. In the second section of this review, we summarize the results of research into the effects of television news coverage on the salience of political issues. The evidence demonstrates that television news engenders an important accessibility bias in the area of political thinking that dictates how people evaluate their politi­ cal leaders. Finally, we move to the subfield of political psychology characterized by the closest ties to cognitive psychology and review emerging strands of research on information processing and political judgment.

ATTRIBUTION-BASED MODELS OF VOTING

The ability of citizens to control the actions of their elected representatives gener­ ally is regarded as the definition of democratic government. In practice, citizens usually have only limited information about the world of public affairs. There­ fore, conventional democratic theory assumes that citizens exercise their influence by approving or rejecting the performance of incumbents, rather than by select­ ing between the platforms and promises of the candidates or parties (Brody & Page, 1972; Fiorina, 1981; Key, 1966). According to this theory of retrospec­ tive voting, the fear of punishment at the polls is what compels politicians to be responsive to their constituents’ preferences. Empirical tests of the theory of retrospective voting have focused primarily on the state of the national economy as the criterion for evaluating the perform­ ance of incumbents. A vast literature has emerged concerning the impact of fluc­ tuations in unemployment, inflation, gross national product (GNP), real income, and other economic indicators on electoral outcomes. With few exceptions, these studies have indicated that periods of economic decline spell defeat for incum­ bents, whereas prosperity leads to reelection (Hibbs, 1987; Kiewiet & Rivers, 1984; MacKuen, 1983). Between elections, the state of national economic indi­ cators is also a harbinger of public evaluations of the incumbent president or governing party. The popularity of incumbents peaks during good times and plunges during bad times (Kinder, 1981; Ostrom & Simon, 1989; Peffley & Wil­ liams, 1985).1 *The attributional account of voting can be generalized across issue areas. However, most of the work in political science demonstrates that economic outcomes are the most important determinants of electoral choices.

148

IYENGAR AND OTTATI

Although the impact of economic factors on voting generally is taken as axio­ matic, research into the psychological processes by which economic conditions influence the thinking of ordinary citizens has lagged behind. Implicit in the aggregate-level analyses of economic voting is a simple attributional model of choice, in which the voter first perceives that economic conditions are below some normative standard and, realizing that incumbent officials are responsible for the state of the economy, decides to vote for challengers. Thus, some studies have shown that when political events are clearly beyond the control of incumbents (such as the OPEC-induced increases in the price of crude oil in the late 1970s), voters appear to take the circumstantial context into account and adjust their evalu­ ations of incumbents accordingly (Hibbs, 1987). Of course, aggregate-level analyses of voting are agnostic as to processes of choice. To pursue the logic of economic voting, individual-level research has focused on two sets of questions. First, what standards do voters employ in as­ sessing economic conditions? In particular, do voters respond to changes in their personal financial circumstances or to changes in national economic conditions? If it is the latter, which particular indicators carry the most weight? In addition to questions of egocentric versus “sociotropic” rationales and indicator diagnosticity, some researchers have attempted to determine whether voters are retrospec­ tive or prospective in orientation. Is past economic performance considered more relevant than future prospects? The second set of issues surrounding economic voting addresses attribution and choice. Here, most research has addressed how voters assign responsibility for economic outcomes (both personal and societal) and the extent to which their attributions influence voting preferences. In addition, some researchers have be­ gun to investigate the importance of information and uncertainty in predicting the public’s assessments of economic conditions and voting choices. Assessing Economic Outcomes

The traditional view of the American voter (Campbell, Gurin, & Miller, 1954) was that pocketbook matters were paramount. The size of John Doe’s paycheck dictated his view of politicians. In contrast, more recent analyses have painted a quite different picture of the electorate. These data suggest that voters are col­ lectivity oriented (i.e., they respond more to the state of the country than their own income). On balance, the evidence seems to indicate that sociotropic judg­ ments exert significantly stronger effects on evaluations of candidates than as­ sessments of personal finances (Kiewiet, 1983; Kinder & Kiewiet, 1979; Weatherford, 1983). Within the literature on sociotropic voting, there are several areas of debate. First, some analysts have pointed out that the dominance of sociotropic over egocentric voting may be an artifact of survey methodology. The standard sur­ vey question used to gauge the egocentric component of voting (whether the voter’s

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

149

own financial situation has improved, stayed the same, or worsened over the past year) does not require respondents to differentiate between changes that can be attributed to governmental action (e.g., a reduction in the prime rate resulting in lower mortgage rates), changes that can be attributed to personal effort (e.g., earning a large year-end bonus), and changes that are more or less random (e.g., winning the state lottery). When analysts have isolated the politically relevant subset of perceived changes in personal finances (either by asking additional ques­ tions or using more sophisticated measurement models), pocketbook voting turns out to be much stronger (Markus, 1988, Rivers, n.d.). An additional area of controversy concerns the relative weight that voters accord assessments of past economic conditions as opposed to expectations regarding future economic conditions. Based on the assumption that voters possess only limited information, most researchers have stressed the importance of the past over the future (Fiorina, 1981). However, there is considerable evidence, both in the United States and in Europe, that voters do respond to perceptions of what the candidates and parties will do in the future (Lewis-Beck, 1988). Little work has been done regarding the conditions that facilitate retrospective or prospective voting. If the sociotropic model of voting is basically valid, the question then arises as to how voters assess the economy? The typical aggregate-level analysis of eco­ nomic voting includes a variety of indicators, such as the unemployment rate, the consumer price index, and changes in real disposable income. The relative importance of these indicators is far from settled. Although some analysts have found that voters particularly are influenced by unemployment data, others have demonstrated strong effects of inflation. Within the domain of inflation, the ques­ tion is whether the consumer price index or more narrowly focused indicators (such as food prices) drive voter responses (Norpoth, Lewis-Beck, & Lafay, 1991). There has been additional debate over the validity of “naive” models of economic voting, in which voters respond directly to indicators of economic outcomes such as the CPI, and “sophisticated” models, in which voters, realizing that economic policies dictate outcomes, respond to indicators of economic policy (Chappell & Keech, 1990). Finally, there has been diverging evidence concerning the form of the rela­ tionship between economic outcomes and political evaluations. Some research­ ers have detected a negative slant to economic voting, by which the political costs of economic hardship exceed the political gains of prosperity (Bloom & Price, 1975; Kernell, 1977; Lau, 1982, 1985). Others have found no traces of such a negativity bias (Fiorina & Shepsle, 1989; Radcliff, 1986). Attributions of Responsibility

As discussed in the preceding section, voters tend to be more sociotropic than egocentric. One explanation—and perhaps the most plausible—for this phenome­ non is that voters do not attribute responsibility for economic outcomes to the

150

IYENGAR AND OTTATI

government. The dominant cultural norms of individualism and self-reliance ef­ fectively depoliticize personal economic adversity. When Americans are asked to explain the changes in their personal economic circumstances, they typically engage in self-analysis, attributing responsibility to themselves (Sniderman & Bro­ dy, 1978). A recent national survey showed that 54% of the explanations offered for changes in personal economic well-being dealt with private behavior, whereas only 12% cited the actions of government (Norpoth et al., 1991). Even the un­ employed, who might have strong motivational grounds to cite government or societal factors, fail to do so (Schlozman & Verba, 1979). Not surprisingly, voters who do attribute changes in their personal economic conditions to governmental or societal factors are much more pocketbook con­ scious in their voting preferences (Norpoth et al., 1991). However, the same result is not obtained when it is the national economy that people are asked to explain. Voters who attribute responsibility for the national economy to the government and those who attribute responsibility to various nongovernment factors (e.g., the actions of labor unions and the productivity of American workers) vote against incumbents during periods of economic adversity. Seemingly, as a practical matter, ordinary citizens are incapable of or unwilling to differentiate between outcomes that are attributable to the actions of government and those that are not. In fact, the typical “attribution” for national economic problems tends to be no more than a restatement of the problem. For instance, a major reason given for unemploy­ ment is that “the economy is in bad shape” (Kinder & Mebane, 1983). Notwithstanding the obvious importance of obtaining a better understanding of the impact of attributions on economic voting, surprisingly little attention has been paid to understanding the origins or development of voters’ attributions. Studies evaluating the effect of various categories of individual differences have shown that there is some evidence of wishful thinking. For example, the attribu­ tions of upwardly mobile people tend to be more dispositional than those of peo­ ple who are less fortunate. Research into the contextual antecedents of attribution had focused on the effects of television news coverage. This research shows that “thematic” framing of news stories encourages the attribution of responsibility for political issues to general societal factors, whereas “episodic” framing en­ courages attribution to private, rather than societal, forces.2 However, the par­ ticular case of unemployment deviated from this pattern. In two separate studies, subjects who watched episodic news reports that focused on individuals who had lost their jobs and subjects who watched thematic reports on national economic trends were equally likely to attribute responsibility (both causal and treatment) for unemployment primarily to societal factors (Iyengar, 1991). 2The essential difference between episodic and thematic coverage is that the former depicts con­ crete events that illustrate issues (e.g., a homeless person in the case of poverty), whereas the latter depicts issues in general terms (e.g., a congressional debate over poverty policy). Between 1981 and 1986, network news coverage of poverty, racial inequality, crime, and terrorism was cast primarily in episodic terms. Contrary to this pattern, news coverage of unemployment was heavily thematic.

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

151

Uncertainty and Economic Voting

Under the typical aggregate-level analysis of economic voting, voters are ration­ al actors. They have well-formed preferences and choose between candidates or parties to maximize personal utility. It is assumed that personal utilities depend, among other things, on a number of economic indicators: interest rates, stocks, inflation, unemployment, growth, and so on. In the real world, the assessment of economic utility is complicated considerably by uncertainty. Lacking familiarity with abstract economic concepts (such as GDP or the CPI), most voters are un­ able to assess with confidence the meaning of significant economic indicators. Moreover, national economic conditions are volatile and susceptible to unpre­ dictable fluctuations. Voters must project past and current economic trends into at least the near future. Unexpected national or international events make those forecasts inaccurate. The uncertainty surrounding voters’ assessments of current and future eco­ nomic conditions has obvious implications for models of economic voting. Most important, voters’ attitudes toward risk (including the risk that they have misassessed economic conditions) enter into determinations of personal utility and, hence, into evaluations of political leaders. People’s taste (or distaste) for uncertainty effects their calculation of utilities. To the extent that a voter dislikes uncertainty, he or she will avoid lesser known candidates. Conversely, voters who seek risk may prefer a lesser known challenger to a better known incumbent, all other things being equal. In short, attitudes toward risk are important characteristics of voters’ comparisons of candidates. (For em­ pirical evidence and further elaboration, see Brady & Ansolabehere, 1989).3 It is well established that most Americans lack the necessary information to compute unbiased estimates of personal finances or national economic conditions. Moreover, most people depend on the mass media for economic information (see the following section) and media reports may provide insufficient or inaccurate 3In the case of electoral choice, attitudes toward risk can be expressed by comparing the expect­ ed utility of economic performance to the utility of the certainty equivalent. The latter is the utility that the individual would ascribe to his or her best prediction of economic conditions if it were cer­ tain that those conditions would in fact materialize. The expected utility of economic performance is the utility that the individual expects to receive as an empirical matter when he or she considers economic conditions. When calculating his or her expected utility, an individual first assigns utilities to alternative economic conditions and then averages these utilities by weighing each alternative con­ dition by its probability of occurrence. Instead of requiring voters to engage in such systematic comparisons between their expected util­ ity of economic performance and the utility they assign to the certainty equivalent, economists use a mathematically equivalent, but simpler, method to represent attitudes toward risk. They classify attitudes toward risk depending on the shape of the utility function. The utility function of the riskaverse voter is concave (i.e., has a positive first derivative and a negative second derivative). The utility function of the risk-seeking voter is convex (i.e., has a positive first derivative and a positive second derivative). Finally, the utility function of a risk-neutral voter is linear. In fact, for any gam­ ble, the risk-averse voter always will assign greater utility to a sure outcome than to a risky alternative.

152

IYENGAR AND OTTATI

information about the state of economic conditions. Therefore, it is likely that voters’ attitudes toward risk will depend on the manner in which economic infor­ mation is presented or framed. In particular, attitudes toward risk may depend on whether outcomes are presented as probable losses or gains. For example, current unemployment figures might reflect an improvement over the course of the past year, but a deterioration with respect to the preceding 5-year average. Thus, the meaning ascribed to objective indicators of economic performance can be influenced by the context in which the evidence is presented. Psychologists have demonstrated that, when offered the choice between a cer­ tain gain and a gamble, people prefer the former; but when the identical choice is phrased in terms of a certain loss or gamble, they prefer the latter. In short, people are risk averse when gambles are framed as gains, but are risk seeking when gambles are framed as losses. (For laboratory evidence, see Kahneman & Tversky, 1982, 1984, 1987; more naturalistic evidence is provided in McNeil, Parker, Sox, & Tversky, 1982; Payne, Laughunn, & Crum, 1980; Thaler, 1980, 1987.) The effects of framing violate the invariance principle of economic theory — the assumption that preference orders for outcomes are unaffected by the manner in which the outcomes are described (for a discussion of the degree to which the invariance assumption is met, see Grether & Plott, 1979; Kahneman & Tversky, 1984). The crucial violation of invariance seems to be that people systematically miscalculate the expected value of gambles (Kahneman, Slovic, & Tversky, 1982). In particular, they overestimate the probability of outcomes that occur with moder­ ate frequency and underestimate the probability of outcomes that occur with very high or low frequency (Kahneman & Tversky, 1984; Quattrone & Tversky, 1988). Generalizing to the case of economic voting, framing will induce individuals to assign a higher probability to changes within the moderate range of economic indicators and a lower probability to changes at the extremes. As a result, news of changes within the moderate ranges will tend to produce larger shifts in as­ sessments of economic performance and the resulting utilities than news of changes at the extremes. Graphically, the result is that the slope of the utilities flattens as the economy is depicted as extremely good or bad relative to the baseline. Therefore, the utility function is concave when the economy is expected to do well and convex when it is expected to do poorly. There has been little evidence concerning the effects of attitudes toward risk on economic voting. Most studies implicitly have treated voters as indifferent to risk by using statistical models in which specific independent variables have linear effects on vote choice (Fiorina, 1981; Kiewiet, 1983). A few studies have incorporated nonlinear terms to capture curvature in the utility functions of voters. The thrust of this literature has been to see if deteriorations in the health of the economy affect incumbents’ approval ratings more than improvements. The evi­ dence has suggested that negative voting is generated by risk aversion (Ansolabehere, 1988; Bartels, 1985; Lau, 1985).

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

153

Quite apart from the literature on negativity, economic research on voting be­ havior has used a standard econometric technique to measure voters’ attitudes toward risk. Citizens are assumed to form preferences about candidates on the basis of some indicator, say, changes in their own income. The indicator is relat­ ed to voting preferences through a utility function, and the statistical problem is to approximate the shape of that function. Tobin (1958) showed that the indi­ cator and the square of the indicator can be used as independent variables in a regression to approximate the shape of the utilities. What is more, the sign of the coefficient on the square of the indicator measures attitudes toward risk. If that coefficient is positive, the individual is risk seeking; if the coefficient is nega­ tive, the individual is risk averse; and if the coefficient is zero, the individual is risk neutral. Tobin’s technique has been applied in studies of both issue voting and economic voting. Enelow and Hinich (1984), Bartels (1985), and Brady and Ansolabehere (1989) all found that voters are consistently risk averse in primary and general elections. In all these studies, attitudes toward risk were assumed to be constant. Thus, questions of framing were considered to be irrelevant. Two experimental studies investigated the validity of this assumption by assessing the effects of framing on voting decisions (Ansolabehere, Iyengar, & Simon, 1990; Quattrone & Tversky, 1988). Quattrone and Tversky administered a series of experiments, in which people were given information about the economy or a public policy and asked to choose a candidate or party. Political choices in these experiments were couched as lotteries and gambles. For example, in one study, subjects were told that party A would produce inflation of 12%, whereas party B would produce inflation of 0% with a probability of .5 and inflation of 24% with a probability of .5. In all cases, Quattrone and Tversky found strong evidence that voting de­ cisions were influenced by framing. When the inflation figures represented a de­ terioration from the past (a relative loss), subjects opted for the riskier of the two candidates and vice versa. Ansolabehere et al. (1990) examined the vote choices of (nonstudent) subjects in a more realistic, but less well-controlled, experiment. In this experiment, sub­ jects were shown a 15-minute clip of a news program into which a 2-minute treat­ ment story had been inserted. The treatments consisted of stories that described the economy as either bad or good (in the present), and as about to get better or worse. The objective was to locate people in the domain of losses or gains and then measure the shape of their utilities for equivalently defined good and bad news. The results indicated that when people used retrospective evaluations of the economy to judge the incumbent president, they were strictly risk averse. However, when they used prospective evaluations, they were susceptible to sig­ nificant framing effects. In sum, research on uncertainty in economic voting has centered on measur­ ing attitudes toward risk. The basic assumption is that voters tend to be risk averse. Citizens’ attitudes toward risk seem to be affected by framing, particularly when

154

IYENGAR AND OTTATI

voters think about the future, rather than the past. To this point, voters’ attitudes toward risk remains uncharted territory. Obviously, more work needs to be done on the malleability of these attitudes if political psychologists are to understand the dynamics of economic voting.

SALIENCE EFFECTS IN PUBLIC OPINION

As a practical matter, most people are unable or unwilling to monitor the course of public affairs carefully. Instead, they attend selectively to a few issues that, for some reason, seem important; and these issues become the principal dimen­ sions for evaluating and judging candidates or parties (Iyengar, 1990a, 1990b; Krosnick, 1988). Because public opinion is based on narrow and issue-specific considerations, changes in the salience of issues are likely to shift the distribu­ tion of preferences, and thereby alter political outcomes. For example, the can­ didate whose positions are most preferred when voters consider the issues of poverty and tax reform might not be the candidate of choice when crime and tax reform are substituted as the focal issues. Accordingly, a great deal of political rhetoric and campaign strategy consists of efforts to capitalize on this “disequilibri­ um of tastes” (Riker, 1986). By bringing forward particular issues at the expense of others, candidates prefer to be evaluated on the issues that place them at a relative advantage (Petrocik, 1990). In contemporary American politics, the mass media—and especially televi­ sion news—are vital components of the process by which issues achieve public salience. For most Americans, the world of public affairs exists primarily in the imagery of television news shows. In this section, we summarize research into the effects of television news coverage on the relative accessibility of political issues. In general, the evidence has indicated that the issues made more accessi­ ble by television news coverage are accorded higher priority by viewers and be­ come especially influential as criteria for choosing between political candidates. Accessibility effects appear primarily in the weights that individuals assign to various considerations when expressing attitudes or making choices. Consider­ ations that are relatively more accessible in memory exert significantly greater effects on attitudes and choices than equally relevant, but less accessible, con­ siderations (Higgins & King, 1981; Wyer & Srull, 1986). Well-known manifesta­ tions of this differential weighting include the tendency to overestimate the frequency of sensationalized events (such as fires and traffic accidents) as causes of death and to underestimate the frequency of quiet risks such as heart disease and stroke (Slovic, Fischhoff, & Lichtenstein, 1982). Presumably, the former come to mind more easily because of the greater publicity and news coverage they attract. Parallel results have been obtained with respect to interpersonal im­ pressions; people typically evaluate their friends or colleagues according to traits or features that are momentarily prominent (Bargh, 1985; Higgins & King,

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

155

1981; Wyer & Hartwick, 1980). Researchers also have shown that attitudes, like information, may be made more or less accessible, and that the more accessible the attitude, the higher the degree of attitude-behavior consistency (Fazio, 1990; Fazio & Williams, 1986).4 In the world of politics, where people must rely heavily on the media for in­ formation, it goes without saying that patterns of news coverage are critical con­ textual determinants of accessibility. Typically, what comes to mind when the citizen thinks about public affairs are the images and information that flash across the television screen. In the discussion that follows, we describe how this type of accessibility bias affects the status of issues in American politics. Television News and Issue Salience

Issues enter and leave the center stage of American politics with remarkable speed. When President Bush was first elected, the problem of illegal drug usage was foremost in Americans’ minds, and the new administration was prompted to an­ nounce a major initiative to deal with this problem. However, a year later, ille­ gal drug usage was mentioned by too few Americans to even warrant inclusion in pollsters’ lists of major problems facing the country. The most plausible ex­ planation of such dramatic shifts in political priorities is couched in terms of pat­ terns of news coverage. In particular, the amount of news coverage accorded political issues is thought to dictate the degree of issue salience. This argument is referred to as media agenda setting. Early agenda-setting studies (conducted in the 1960s) were plagued by a num­ ber of methodological difficulties, including confusion between cause and effect. Did the convergence of newspaper readers’ political concerns and newspaper con­ tent mean that news coverage had set the audience agenda, or did it mean that editors and journalists had tailored their news coverage to appeal to the political concerns of their readers? In response to such ambiguities, communications researchers began to track the evolution of public concern for particular issues and events in relation to changes in the pattern of news coverage. With few ex­ ceptions, these time-series studies uncovered evidence of significant media agendasetting effects (Rogers & Dearing, 1988). 4Parallel accessibility effects have been detected in studies of survey responses. Public opinion researchers have demonstrated that the wording, format, and ordering of questions produce dramatic variations in reported beliefs or opinions. Thus, people describe themselves as disinterested in poli­ tics if they are first asked a series of difficult factual questions concerning the identity and activities of various public officials. On the other hand, if they are asked about their political interest before being confronted with the factual knowledge questions, they describe themselves as substantially in­ terested (Bishop, Oldendick, & Tuchfarber, 1982). Similarly, the percentage of respondents favor­ ing more generous federal financial assistance is markedly higher if the recipients of such assistance are described as poor people, rather than people on welfare (Smith, 1987; for a general review of accessibility effects in surveys, see Zaller & Feldman, 1988).

156

IYENGAR AND OTTATI

The time-series analyses further refined the agenda-setting paradigm by in­ corporating indicators of events (i.e., real-world cues) that might be experienced directly in addition to the level of media coverage as potential determinants of the public agenda. When events were relatively obtrusive (e.g., energy short­ ages, rising prices, or widespread crime), the evidence indicated that they affect­ ed public opinion directly. For instance, when the level of unemployment increased significantly, more people mentioned unemployment as a major national problem independently of how much news coverage the media provided (Behr & Iyengar, 1985; MacKuen, 1981). In addition to the course of events, the level of presidential rhetoric was found also to influence the public’s issue agenda. When the presi­ dent addressed the nation on a particular problem and the address was televised nationwide, he was able to boost public concern independently of the amount of other news coverage accorded that problem (Behr & Iyengar, 1985). Finally, in a further elaboration of the interrelationships between events, network news, and public opinion, Behr and Iyengar (1985) demonstrated that agenda setting generally was unidirectional—news coverage affected the level of public concern, but public concern did not affect the focus of television news. Instead, television news coverage was determined primarily by objective events. Agenda-setting effects have been captured for all forms of mass media cover­ age in experimental and survey-based studies; with open-ended indicators, in which respondents identified the “most important problems facing the country”; and with closed-ended items, in which they rated the importance of particular issues. These effects have been observed for both local and national problems. In all these areas, research has shown that individuals habitually refer to issues or events “in the news” when diagnosing current social and political ills (Iyengar & Kinder, 1987). Of course, the media agenda is not adopted uniformly by all members of an audience. The progressive refining of agenda-setting research has included ef­ forts to identify segments of an audience that are likely to be more or less vulner­ able to agenda setting. In Iyengar and Kinder’s (1987) experiments, individuals with lower levels of education, political interest, and political participation were especially susceptible to influence. The less educated, interested, and involved segments of the audience presumably are less able to retrieve independently de­ rived information that might cast doubt on the message contained in news reports. The availability of information that permits critical analysis (counterarguing) of media presentations is an important factor affecting ability to resist media influence, either in the form of agenda setting or persuasion (Petty, Ostrom, & Brock, 1981). Erbring, Goldenberg, and Miller (1980) were the first to suggest that individuals differed in their receptivity to news about particular issues. Elderly people may be more attentive to news about crime, whereas people working for defense con­ tractors may be first to respond to news coverage of impending layoffs and un­ employment. Iyengar and Kinder’s experiments confirmed that news coverage and the personal circumstances of the audience interactively shape perceptions of national issues. Retired viewers were more likely than younger viewers to

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

1 57

mention social security as a pressing problem following exposure to news reports on social security, and Blacks cited racial discrimination as a significant problem more frequently than Whites after watching news coverage of the issue (Iyengar & Kinder, 1987; for additional evidence concerning the joint effects of news cover­ age and the personal relevance of news reports see Tyler, 1980; Zucker, 1978). Television News and Differential Weighting of Issues in Candidate Evaluation

In addition to affecting the sheer salience of issues, television news coverage also influences the degree to which issues are used as criteria to evaluate political lead­ ers. This priming effect concerns the impact of news coverage on the weight in­ dividuals assign to their opinions on particular issues when they make political evaluations.5 In general, the more prominent an issue is in the national infor­ mation stream, the greater the impact of opinions regarding that issue on politi­ cal evaluations. In the context of election campaigns, priming means that an individual voter’s choice between political candidates is likely to be based on an analysis that pro­ vides greater weight to preferences on issues that receive heavy news coverage than to preferences on less newsworthy issues. For example, in a study of the 1982 election in the third congressional district of Connecticut, researchers found that voters who felt more optimistic about national economic conditions were more supportive of the Republican incumbent. However, when exposed to news cover­ age of the economy, the effect of voters’ optimism on evaluations of the candi­ dates was more than tripled (Iyengar & Kinder, 1987). An even stronger priming effect emerged with respect to participants’ perceptions of the personal traits of the candidates. Viewers generally preferred the candidate in whom they saw more positive personal characteristics. But following exposure to news reports about the candidates’ personal backgrounds, the weight accorded to trait ratings was increased nearly fivefold (Iyengar & Kinder, 1987). Priming effects also have been detected in nonexperimental studies. Using na­ tional survey data, Krosnick and Kinder (1990) found that the public’s support for U.S. intervention in Central America became twice as influential as a deter­ minant of President Reagan’s popularity in the period immediately following the disclosure that funds from the sale of arms to Iran had been used to finance the Nicaraguan Contras. Similarly, public evaluations of presidential performance became significantly more responsive to foreign policy beliefs in the aftermath of the Iraqi invasion of Kuwait (Iyengar & Simon, 1992). Media priming effects are especially important in the context of primary elec­ tion campaigns. Recent analyses have centered specifically on the effects of “horse 5As used by communication researchers, priming refers to simply enhancing the accessibility of a criterion that already is available (Iyengar & Kinder, 1986).

158

IYENGAR AND OTTATI

race” coverage in the making and unmaking of American presidential candidates. As countless studies of campaign journalism have shown, stories detailing the candidates’ electoral prospects—their poll standings, delegate counts, fund rais­ ing efforts, and related campaign indicators—have become the staple of campaign reporting and frequently dwarf coverage of more relevant facets of the campaign (Robinson & Sheehan, 1980). The prominence of viability in the news stream virtually guarantees that perceptions of the candidates’ electoral viability will pro­ vide a strong evaluative impetus. Bartels (1988) demonstrated that voters prefer the candidate who is deemed more viable, and that voters with positive feelings toward a particular candidate are especially likely to vote accordingly if they con­ sider the candidate viable. Naturally, the magnitude of media priming effects depends on the content of news coverage. Most importantly, these effects are magnified when news reports convey explicit cues about governmental responsibility for the state of national problems. For example, in Iyengar and Kinder’s (1987) experiments, the weight accorded economic performance in overall evaluations of President Reagan in­ creased when subjects were given information about economic problems, and in­ creased further when individuals watched news reports on economic problems suggesting that Reagonomics was the principal cause of these problems. The medi­ ating effects of presidential responsibility in strengthening the priming effect were detected for news coverage of both political accomplishments and failures. For example, overall evaluations of President Carter were equally primed (i.e., equally influenced by assessments of Carter’s performance in foreign affairs) by news coverage of the Carter administration’s major foreign policy debacle—the Irani­ an hostage crisis—and by news coverage of the administration’s major foreign policy success—the Camp David accords. In short, the issues and events portrayed in television news programs become the issues and events that color voters’ attitudes toward public officials and polit­ ical candidates. Heavy news coverage of a particular issue makes that issue par­ ticularly salient and more likely to be used as a criterion for political evaluation.

MODELS OF INFORMATION PROCESSING

Models that simply delineate the situational (e.g., state of the economy) or in­ dividual difference variables (e.g., respondent socioeconomic status) that influence political judgment and behavior generally fail to provide an adequate understanding of the determinants of these decisions. In the previous sections of this chapter, we focused on certain isolated components of the psychological process that in­ tervene between exposure to political information and judgment (e.g., attribu­ tion, accessibility). In this section, we attempt to place these and other politically relevant phenomena within the context of an information processing model of political judgment and decision making.

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

159

Typically, social information-processing models delineate the stages of infor­ mation transformation that occur en route to the report of an overt judgment. The first two stages are exposure and comprehension. We begin by elaborating on these two initial stages. We then proceed to discuss later stages of processing that occur en route to formulating a political judgment. These include encoding and interpretation, organization, storage, retrieval, and integration. As is soon apparent, these later stages of processing differ considerably, depending on the judgmental objective the citizen has in mind. No individual is exposed to the full universe of information available in the political environment. To the contrary, individuals are exposed to only a subset of all the political information that is objectively present. Hence, a central ques­ tion is what determines the range and nature of the political information that each individual encounters. We make a distinction between two mechanisms of selec­ tive exposure—involuntary and voluntary (Wyer & Ottati, 1992). In the case of the former, individuals have little voluntary control over the political informa­ tion they encounter. By giving coverage to certain news stories while ignoring others, the mass media exert substantial control over what information is presented selectively to the individual (i.e., gatekeeping effects). Similarly, interpersonal sources of political communication, such as parents and friends, also may in­ fluence exposure by selectively presenting the individual with a limited or biased sample of political ideas and viewpoints. However, it is also true that individuals often play an active and voluntary role in seeking and avoiding exposure to certain political environments and mes­ sages. Individuals exercise this sort of behavioral control over information ex­ posure in a variety of ways, ranging from selecting a particular television channel or newspaper article to selecting acquaintances and friends for affiliation. Even when two individuals attend to the same stimulus, they may differ in terms of the amount of attention allocated to a given stimulus dimension (Smith, 1990; Smith & Zarate, 1992). For example, perceivers who are racially prejudiced may pay more attention to a candidate’s skin color. The individual’s processing goal also may influence the extent to which attention is allocated to a politician’s in­ dividuating attributes or, alternatively, to only stereotype-related attributes (e.g., skin color). Typically, individuals will seek out and attend to information that is consistent with their prior knowledge and expectations (Snyder & Swann, 1978; Wyer, Strack, & Furhman, 1984). Involuntary and voluntary mechanisms of selective exposure sometimes oper­ ate in tandem to produce strong adult predispositions to judge political events and actors in accordance with childhood predispositions. By virtue of birth, an individual may be exposed involuntarily to predominantly conservative political opinions and beliefs. That same individual subsequently may seek out conserva­ tive messages and actively avoid liberal messages that are inconsistent with this prior disposition. The conservative bias associated with what information enters the individual’s information processing system should, in turn, be reflected in

160

IYENGAR AND OTTATI

a conservative bias associated with the judgments that are output from this in­ dividual’s cognitive system. Even when exposed to the same political message, individuals may differ in terms of their capability to comprehend the message. The second act of informa­ tion processing is to comprehend the political message in terms of low level noun, attribute, and action concepts drawn from long-term memory. This comprehen­ sion or recognition process is analogous to looking up words in a dictionary where the word is simply the stimulus configuration in its raw sensory form and the definition is the concept stored in long-term memory that reflects the meaning of the word. This analogy is not restricted to written or verbally presented mes­ sages. The word may be a visual image (e.g., facial expression) that is compre­ hended by matching it to an appropriate concept in long-term memory (e.g., an angry face). A detailed discussion of the comprehension process is beyond the scope of this chapter (for a review, see Lachman, Lachman, & Butterfield, 1979). Yet, it is important to recognize that a political perceiver’s mental dictionary varies depending on political expertise (Graber, 1988; Fiske & Kinder, 1981; Fiske, Kinder, & Larter, 1983; Lau & Erber, 1985) and any other characteristic as­ sociated with this domain-specific form of knowledge (e.g., education, socioeco­ nomic status). As a consequence, political sophisticates are more likely to comprehend political messages than nonsophisticates, and, over time, political sophisticates increase their political knowledge base at a faster rate (Tichenor, Donohue, & Olien, 1970, 1980). Once a political message is comprehended, the comprehended message must be processed further to arrive at a political judgment. As noted previously, we consider five additional stages of processing that mediate the report of an overt judgment: encoding and interpretation, organization, storage, retrieval, and in­ tegration. When discussing these stages of information processing, it is impor­ tant to consider three concerns that come into play at several different stages. First, due to limitations in cognitive capacity and motivation, individuals often are unable or unmotivated to deliberatively process all the information that is relevant to a judgment or decision. Instead, heuristics are employed at a number of stages of information processing. Second, judgments and decisions are not de­ termined exclusively by the semantic content of acquired political information. To the contrary, episodic affective responses experienced at any stage of pro­ cessing can play an important role in determining the final judgment or deci­ sion. Third, and perhaps most importantly, the manner in which information is processed depends on the purpose for which an individual expects to use the in­ formation. For example, a politician’s speech concerning the Middle East con­ flict may be processed one way if the individual’s goal is to form a policy preference (e.g., Should the United States support Israel?), but quite another way if the per­ ceiver’s goal is to evaluate the politician’s suitability for public office (Wyer & Ottati, 1992).

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

161

During campaigns, the objective of evaluating or forming an impression of a candidate is particularly important. Between campaigns, citizens are more likely to possess the objective of forming an opinion about some political issue (Lang & Lang, 1959). The manner in which an individual encodes, interprets, organizes, stores, and combines information to arrive at a judgment depends on which of these objectives is paramount. Thus, our discussion of the remaining stages of political information processing is divided into two sections: one concerning processing en route to the formation of an issue opinion, and the other concern­ ing processing en route to evaluating a politician’s suitability for office. Information Processing en Route to Forming Issue Opinions

Encoding and Interpreting Issue Messages. The recipient of a political mes­ sage often is motivated to understand the comprehended information in terms of abstract concepts. Yet, any given message usually can be interpreted in terms of more than one concept. For example, a message indicating that the Ku Klux Klan should be allowed to hold public meetings can be interpreted as either racist or civil libertarian. People typically do not conduct a detailed evaluation of the alternate concepts that potentially might apply to interpreting a given message. Rather, they apply the concept that is most accessible at the time they are ex­ posed to the message (Bargh, 1985; Higgins & King, 1981; Wyer & Srull, 1981, 1986, 1989). A message recipient who wishes to form an opinion about the issue or event to which the speaker refers may try to construct a temporally or causally related scenario of the policy event that includes its antecedents and consequences (McGuire & McGuire, 1991; Pennington & Hastie, 1986). For example, a per­ ceiver may encode a policy proposal in terms of the consequences it poses to the perceiver individually or in terms of those it poses to some social group with which the perceiver identifies. Emphasis on the former is likely to heighten the accessibility of consequences relevant to individual self-interest. Emphasis on the latter is likely to enhance the accessibility of consequences relevant to sociotrop­ ic or group interest. Research in a number of policy domains (e.g., school bus­ ing) has suggested that encoding in terms of group-level consequences often predominates over the former (Kinder & Kiewet, 1979; Sears & Funk, 1990). In many cases, persuasive policy communications consist of a series of argu­ ments that explicitly delineate the positive or negative consequences associated with a proposal (Lau, Smith, & Fiske, 1991). A central question involves the extent to which an individual accesses prior relevant information when encoding a persuasive policy communication of this nature. Lau et al. presented subjects with a policy communication that emphasized either one or two distinct sets of consequences. For example, to fight overcrowded prisons, one set of arguments emphasized that private companies could build and operate prisons more effi­ ciently in a cost-effective manner (efficiency interpretation). Another set of

162

IYENGAR AND OTTATI

arguments emphasized that government-run prisons financed by a small increase in income tax would protect better the rights and safety of prisons, guards, and the surrounding community (protection interpretation). When the proposal was accompanied with only one of these interpretations, it was encoded in terms of the presented interpretation (efficiency or protection), and prior knowledge ex­ erted no influence on encoding. When presented with both sets of arguments, the proposal was encoded in terms of whichever interpretation was more chroni­ cally accessible to the voter. Thus, in this latter case, prior knowledge played a significant role in determining the nature of the encoding and interpretation process. The relative accessibility of the two interpretations (efficiency vs. protection) may, in turn, reflect the individual’s priorities with regard to more abstract values. This suggests that specific policy beliefs may be organized around superordinate core values that are chronically accessible to the voter. We return to this possi­ bility presently. Intra-Issue Organization. When considering the organization and represen­ tation of issue information in memory, it is useful to distinguish “intra-issue” or­ ganization from “inter-issue” organization. Intra-issue organization refers to the manner in which an individual organizes the specific antecedents and consequences associated with a given policy event or proposal. Inter-issue organization refers to a more overarching organizational structure that maps the linkages between various issue beliefs. The question of how issues are structured internally when represented in long­ term memory remains virtually unanswered by students of political opinion. A conceptualization of thought systems recently proposed by McGuire and McGuire (1991), although yet to be fully explored by students of political cognition, may offer a framework for understanding the structure of policy representations. McGuire and McGuire suggested that cognitions closely associated with a policy event (e.g., the U.S. bombing of Iraq) can be distinguished in terms of a 2 x 2 x 2 taxonomy. The dimensions of this factorial are antecedents versus conse­ quences, desirability versus undesirability, and promotive versus preventive relevance to the policy event. Thus, it is possible to distinguish four types of an­ tecedents: desirable antecedents that make the event more likely, undesirable an­ tecedents that make the event more likely, desirable antecedents that make the event less likely, and undesirable antecedents that make the event less likely. McGuire and McGuire distinguished four analogous consequences: desirable consequences that are promoted by the policy event, undesirable consequences promoted by the policy event, desirable consequences prevented by the policy event, and undesirable consequences prevented by the policy event. From a normative point of view, a person’s policy opinion (e.g., Should the United States bomb Iraq?) should be based on the expected utility of the associated consequences.

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

163

McGraw and Pinney (1990) focused on the cognitive structure of consequences associated with a specific policy proposal, namely, the 1986 Tax Reform Act. They examined the clustering of these cognitions using the adjusted ratio of cluster­ ing index (Roenker, Thompson, & Brown, 1971) and found considerable evi­ dence of evaluative clustering. That is, positive and negative consequences associated with the 1986 Tax Reform Act were represented in separate clusters. Furthermore, the tendency for this information to be clustered evaluatively was more pronounced among political sophisticates than among political nonsophisti­ cates (McGraw & Pinney, 1990). Inter-Issue Organization and Representation. The question of inter-issue representational structure has been the focus of considerable debate among polit­ ical psychologists who have sought to delineate the structure of political belief systems. Virtually all of this research has, either implicitly or explicitly, embraced a hierarchical model of inter-issue representation. That is, specific issue beliefs have been hypothesized to be nested under and vertically associated with more central and superordinate crowning postures (Converse, 1964) or core values (Hurwitz & Peffley, 1987). Converse (1964) considered the possibility that all issue beliefs are structured in accordance with a single, overarching liberal/conservative ideological dimen­ sion. He found little evidence for the existence of such a structure. Knowing that an individual adopts a conservative position on one issue (e.g., social security) does not necessarily imply that an individual will adopt a conservative position on another issue (e.g., federal aid to education). Although some have suggested that Converse’s findings are limited to a given historical context (Nie & Ander­ son, 1974) or method of measurement (Sullivan, Pierson, & Marcus, 1978), most researchers have agreed that issues are not represented in accordance with a sin­ gle liberal/conservative ideological posture. Hurwitz and Peffley (1987) suggested that citizens organize issues under mul­ tiple crowning postures. They emphasized that Converse’s (1964) operationali­ zation of issue constraint only focused on the horizontal inter-issue linkages in representation. Studies that measure the vertical linkages between specific issues and crowning postures often have found considerable evidence of constraint (Conover & Feldman, 1984; Lane, 1962; Marcus, Tabb, & Sullivan, 1974; Peffley & Hurwitz, 1985). Furthermore, the crowning postures that serve to organize specific issues may differ across issue domains (e.g., economic, social, foreign) and individuals. Conover and Feldman (1984) argued that in several issue domains there are two distinctly different conservative postures. In other domains, a crowning posture or schema may incorporate both liberal and conservative issue elements. A su­ perordinate posture is rarely bipolar in the liberal/conservative terms. Further­ more, although some citizens may be schematic with regard to a given superordinate posture (e.g., free market conservatism), other citizens may be

164

IYENGAR AND OTTATI

aschematic with regard to this posture, but schematic with regard to another (e.g., Democratic socialism). As a consequence, individuals who are knowledgeable with regard to one issue domain may not be knowledgeable with regard to other issues (Iyengar, 1990a). Judd and Krosnick (1989) proposed a model of political memory organization that incorporates the interrelation between abstract values or ideological princi­ ples, specific policies, political reference groups, and individual political actors. In accordance with a network formulation of memory organization (Anderson, 1983; Collins & Quillian, 1969), these political objects are represented as object nodes. Each object possesses a strength (i.e., level of accessibility or activation) and evaluation (positive vs. negative). Two nodes become linked when they are thought about simultaneously. Information about the implicational relation be­ tween the linked nodes (i.e., positive, favors, implies vs. negative, opposes, or contradicts) is stored with each link. Node linkages vary in terms of the strength of association. Implicit in the Judd and Krosnick (1989) model is the notion that specific is­ sue positions are organized around multiple crowning postures or values (e.g., equality, freedom). Consistent with Tetlock’s (1983, 1986) value pluralism model, a specific policy alternative (e.g., school busing) may be associated positively with one cherished value (e.g., equality), but associated negatively with another equally cherished value (e.g., freedom). Generally speaking, the model assumes that political experts possess political belief systems containing greater consistency among issues and values than do political novices. Yet, to the extent that experts are more sensitive to conflicts of value, Tetlock’s value pluralism model suggests that experts also can have difficulty maintaining a perfectly constrained or con­ sistent political belief system. Retrieving and Combining Information to Arrive at a Political Issue Opin­ ion. People who are required to report a political opinion do not retrieve and use all of the knowledge they have accumulated that potentially bears on this judg­ ment (Kahneman, et al., 1982; Nisbett & Ross, 1980; Taylor & Fiske, 1978; Wyer & Hartwick, 1980). Rather, they access only a subset of information that they consider representative of the pool of knowledge they have acquired, and therefore to be of sufficient basis for making the judgment. When asked to express a policy position (e.g., should the Ku Klux Klan be permitted to speak freely on campus?), one possibility is that individuals retrieve an accessible subset of the available consequences stored in memory. This acces­ sible set of information may include self-generated consequences, as well as con­ sequences that previously were communicated to the individual. This set of consequences then would be combined to arrive at an overall policy position. A variety of piecemeal (Fiske & Pavelchak, 1986) combinatorial rules have been introduced to model this process, ranging from additive (Fishbein & Ajzen, 1975) to weighted average (Anderson, 1965).

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

165

Other possibilities are suggestive of a less deliberative process. Individuals may employ cognitive shortcuts or heuristics when arriving at a policy opinion. For example, one may retrieve the notion that “I hate the Ku Klux Klan,” and therefore express the opinion that it should not be granted the privilege of speak­ ing in public (Kuklinski, Riggle, Ottati, Schwarz, & Wyer, 1991; Ottati, Riggle, Wyer, Schwarz, & Kuklinski, 1989). This possibility entails no consideration of policy consequences whatsoever. Alternatively, retrieval may proceed only so far as accessing one or two superordinate core values. For example, an in­ dividual may retrieve a cherished value of freedom of speech, and thereby en­ dorse the proposition that the Ku Klux Klan should be granted this right. All of the possibilities just mentioned assume that information is brought to bear on a specific issue opinion at the time an individual is asked to report the opinion. In some cases, individuals simply may retrieve a prepackaged policy position that has been formulated previously. According to McGraw and Pinney (1990), attitudes toward a policy option can be formulated on-line. In the on-line model, issue-relevant information is used to form and update an issue opinion during exposure to policy communications. The integrated opinion tally is then stored in long-term memory and later retrieved to report a judgment. Which route to policy judgment actually is taken? It is unlikely that there is any general answer to this question. Rather, individual difference (e.g., need for cognition, political sophistication) and situational factors probably moderate the tendency for citizens to follow one route or another. Information Processing Models of Candidate Evaluation

Traditional models of candidate evaluation and voter decisionmaking failed to delineate explicitly the cognitive processes that underlie these political decisions. The sociological approach (Lazarsfeld et al., 1949) portrayed the vote decision as a response to group affiliations and social pressure. In essence, the vote choice was viewed as a group decision. As such, cognitive processes operating at the individual level were largely ignored. The “Michigan model” (Campbell, Con­ verse, Miller, & Stokes, 1960; Campbell et al., 1954) proposed that the vote decision is determined by attitudes toward the candidates and parties. While in­ corporating the psychological construct of “attitude,” this model was largely si­ lent about the particulars of the mediating cognitive process. Rational choice theorists (e.g., Downs, 1957; Enelow & Hinich, 1984) proposed a normative model of voter decision making that draws heavily on expected utility theory. This approach involves mental computations that place unrealistic demands on the limited cognitive system. In sum, most of the traditional approaches to voting behavior have failed to provide a comprehensive or realistic account of the stages of information process­ ing that mediate the overt expression of a voting preference. In the remainder of this section, we discuss recent voting behavior research that has explicitly

166

IYENGAR AND OTTATI

adopted a cognitive perspective. We divide our presentation into a discussion of encoding and interpretation, organization, storage and representation, and integra­ tion. Encoding and Interpreting Candidate Information. When the recipient of a political message possesses the objective of forming an impression of a candi­ date, the recipient often is motivated to interpret this information in terms of pri­ or knowledge. Yet, as noted earlier, any given message often can be interpreted in terms of more than one concept. In these instances, the most accessible, rele­ vant concept is the one that is applied. In addition to being influenced by the per­ ceiver’s processing objective, concept accessibility is determined by the frequency and recency of prior concept use, as well as previously acquired knowledge about the candidate. Frequency of prior use often is associated with individual differences in chronic accessibility (Anderson, Reynolds, Schallert, & Goetz, 1976; Bargh, Bond, Lom­ bardi, & Tota, 1986; Bruner, 1951; Klinger, 1977). For example, Lodge and Hamill (1986) and Lodge, McGraw, and Stroh (1989) demonstrated that parti­ san schematics are more likely to interpret and encode political messages in terms of abstract party stereotypes. Quite fortuitous situational events also can influence concept accessibility and, as a result, the nature of the encoding or interpretation (Higgins & King, 1981; Wyer & Srull, 1989). For example, the message “Rea­ gan bombs Libya” may be interpreted differently depending on whether this in­ formation immediately follows exposure to a news report about Hitler’s bombing of London or a story concerning the administration’s battle against South Ameri­ can druglords. In this sense, the flow of news stories may influence interpreta­ tions of any given political message indirectly. A second determinant of concept accessibility is previously acquired knowledge about the object to which the information pertains or, alternatively, about the group or category to which the target belongs. Consider again the statement “The Ku Klux Klan should be permitted to speak freely in public.” If this statement is attributed to Jesse Helms, it is likely to be interpreted as racist, whereas, if attributed to Jesse Jackson, this same statement is likely to be interpreted as civil libertarian. Presumably, the racist interpretation is consistent with prior, acces­ sible knowledge about Helms (White, conservative, racist), whereas the civil liber­ tarian interpretation is consistent with prior accessible knowledge about Jackson (Black, liberal, nonracist). Lodge and Hamill (1986) and Lodge et al. (1989) reported analogous findings when considering the impact of prior knowledge about a candidate’s party membership. The prior, accessible concept used to interpret a political communication need not be abstract or generic in nature (e.g., a racial group stereotype). Prior ex­ perience with a particular political figure may influence the interpretation of re­ marks made by a similar, subsequently encountered political figure. For example, an ambiguous remark made by Iraq’s Saddam Hussein may be interpreted as

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

167

dangerous, because he reminds the perceiver of Adolf Hitler, not because of his racial or ethnic origin. Simple attention to Hussein’s moustache may trigger such a process (Smith & Zarate, 1992). How does concept accessibility influence the encoding and interpretation of information about candidates? A variety of effects have been proposed. Accord­ ing to the selective encoding hypothesis, aspects of a political message that are interpreted easily in terms of concepts that are accessible at the time the message is received are likely to be encoded. In contrast, those aspects that cannot be in­ terpreted in terms of accessible concepts often are ignored. Consequently, the particular concepts that happen to be most accessible at the time the message is received determine which aspects of the message are encoded and retained in long-term memory (Lingle, Geva, Ostrom, Leippe, & Baumgardner, 1979; Wyer & Gordon, 1982; Wyer, Srull, Gordon, & Hartwick, 1982). Consistent with this notion, Lodge and Hamill (1986) and Lodge et al. (1989) found that individuals recalled a higher proportion of a candidate’s party consistent issue positions than of a candidate’s party inconsistent positions. Due to selective encoding, candi­ date attributes that are inconsistent with accessible prior expectations evidently are deleted. This discussion pertains to the encoding of political information that is unam­ biguously consistent or inconsistent with the accessible concept. In contrast, am­ biguous messages are, by definition, interpretable in terms of more than one concept. In this case, it is the more accessible concept that is used for interpreta­ tion (Bargh, 1985; Carver, Ganellen, Froming, & Chambers, 1983; Higgins & Chaires, 1980; Higgins & King, 1981; Srull & Wyer, 1979, 1989; Wyer & Srull, 1981). Here, once again, the Ku Klux Klan example is instructive. Presumably, the concepts of racist and civil libertarian are both potentially relevant to inter­ preting the statement “The Ku Klux Klan should be permitted to speak freely in public.” However, it is the relative accessibility of these two concepts at the time of message acquisition that determines which concept will be applied. Insofar as the expression of ambiguous position statements may enhance a candidates electability (Page, 1976; Shepsle, 1972), effects of this type may be especially prominent in candidate perception. Although candidate evaluation research explicitly has demonstrated selective encoding effects (Lodge & Hamill, 1986; Lodge et al., 1989), it has yet to demon­ strate selective interpretation of ambiguous information at the encoding stage of processing. The effects of evaluative consistency biases (balance and agreement) have been considered (Krosnick, 1989; Ottati, Fishbein, & Middlestadt, 1988). Yet, preliminary evidence suggests that these effects do not occur during encod­ ing. Rather, they appear to reflect the influence of cognitive readjustments that occur at a later stage of processing (Ottati, Terkildsen, & Hubbard, 1992). The earlier discussion focuses on how interpretations of ambiguous messages tend to be similar to the activated concept (i.e., an assimilation effect). Yet, in some cases, a contrast effect can emerge (Herr, Sherman, & Fazio, 1983;

168

IYENGAR AND OTTATI

Manis, Nelson, & Shedler, 1988). This is especially likely when the implica­ tions of the acquired message are diametrically opposite to those of the accessi­ ble construct. That is, the activated concept may serve as a standard for comparison when considering the implications of the acquired message (see Martin, Seta, & Crelia, 1990, for an alternative model of this cognitive process). Effects of this nature may be prominent when people interpret a party renegade’s remarks. For example, a conservative position statement may be interpreted as more con­ servative when expressed by a Democrat than when expressed by a Republican. Contrast effects of this nature also may be evoked by the juxtaposition of two candidates competing for office. This is because concepts activated by the mes­ sage of one candidate may serve as anchors when interpreting the statements made by a competing candidate. Thus, in a preconvention context dominated by Jesse Jackson, Dukakis’ statements may have been interpreted as moderate. Yet, in the postconvention context dominated by Bush, similar statements made by Dukakis may have been interpreted as liberal. Political cognition researchers have yet to empirically document contrast effects of this nature. Organization o f Candidate Information. Once a message has been encoded and interpreted in terms of a given concept, it often is thought about in relation to prior knowledge or information. For example, a behavior that is encoded as dishonest may be thought about in relation to other honest or dishonest behaviors previously learned about the candidate. As a result, a knowledge configuration is formed consisting of the encoded piece of information, previously acquired pieces of information, and associational links between them. This knowledge con­ figuration is structured metaphorically as a network (Collins & Loftus, 1975), consisting of concept nodes and interconnected links or paths that vary in terms of strength of interconcept association. If the perceiver’s objective is to form an impression of a candidate, this configuration may consist of a superordinate trait node along with various behaviors and messages that are interconnected. The exact nature of this associational structure has been the subject of con­ siderable debate. In one model, it is assumed that incongruent behaviors (e.g., a dishonest remark encoded into a knowledge configuration that predominantly contains honest behaviors) are more likely to forge interbehavioral associative pathways. According to Hastie (1980) and Srull (1981), this associative activity reflects the perceiver’s attempt to explain why the incongruent behavior occurred. Yet according to Hamilton, Driscoll, and Worth (1989), this tendency does not apply when individuals acquire numerous pieces of behavioral information en­ coded in terms of multiple trait categories. Under this latter condition, Hamilton et al. (1989) argued that all behaviors are associated directly with each other with­ in a given trait-behavior cluster. Wyer and Gordon (1982, 1984), Srull and Wyer (1989), and Wyer (1989) pro­ posed a dual-coding model of organization. In this model, each behavior is as­ sumed to be represented in both a trait-behavior cluster and an evaluation-behavior

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

169

cluster. Behaviors with opposite trait implications are assumed to be organized in separate trait clusters (e.g., an “honest” cluster and a separate “dishonest” cluster), but inconsistent behaviors are organized in a single evaluation-behavior cluster. Within the evaluation-behavior cluster, the accessibility of a given be­ havior is a function of two competing mechanisms. Initially, evaluatively incon­ sistent behaviors may have an accessibility advantage for reasons analogous to those proposed by the Hastie-Srull model. Yet, over time, individuals may review behaviors consistent with the central evaluative concept to confirm its validity. This bolsters the link between the central evaluative concept and evaluatively con­ sistent behaviors. As a result, the relative accessibility of evaluatively consistent behaviors eventually may outweigh that of the evaluatively inconsistent behaviors (Wyer & Martin, 1986). Research on memory for political actors invariably has produced a memory advantage for information that is semantically or evaluatively consistent with more global candidate assessments. Semantic consistency effects are implicated by recog­ nition memory tests for issues presented about a party labeled candidate. That is, memory for party consistent issue stands is superior to memory for party in­ consistent issue stands (Lodge & Hamill, 1986; Lodge et al., 1989). Analysis of recognition data also produces a separate and independent evaluative consistency (balance) effect. Namely, memory for issue stands that are evaluatively consis­ tent with global candidate evaluations is superior to memory for evaluatively in­ consistent issue stands (Lodge et al., 1989). However, all of these studies have employed a recognition measure of memory. Unlike free recall, the recognition task taps inferences or “best guesses” even when a particular issue is not accessible to the respondent. In this research, 75% of the presented issue stands were consistent with the presented party label. Thus, stereotype-based inferences were likely to increase recognition scores for party consistent information, but decrease recognition accuracy scores for party incon­ sistent information. In other words, the consistency effects may not pertain spe­ cifically to issue accessibility. Furthermore, in all this research, subjects performed a distractor task prior to completing the recognition test. Thus, enough time may have passed to eliminate inconsistency effects (Wyer & Martin, 1986). Representation of Candidate Information in Long-Term Memory. We have considered the process whereby citizens organize specific pieces of information (i.e., behaviors, messages) when receiving information about a political candi­ date. As such, we focused on how information is organized within a given knowledge configuration. The knowledge configurations that result from this ini­ tial organizational process presumably are stored in long-term memory as part of a larger representation of the candidate. We now turn to the content and struc­ ture of this larger representation in long-term memory. In many cases, political cognition researchers have proposed a hierarchical network model of candidate representation. According to McGraw, Pinney, and

170

IYENGAR AND OTTATI

Neumann (1991), a superordinate person node is linked vertically to two infor­ mation clusters. One contains the candidate’s issue positions, whereas the other contains traits. The issue cluster may be organized as previously noted when describing the structure of belief systems, but this has yet to be considered when focusing specifically on the organization of a candidate’s issue positions. It com­ monly is assumed that candidate trait information is not organized simply as a multitude of traits. Rather, people are assumed to organize trait information in terms of a limited number of broad trait categories (Kinder, 1986; Miller, Wattenberg, & Malanchuk, 1986). Furthermore, it has been argued that these categories remain relatively constant across elections (Miller et al., 1986). For example, Kinder (1986) proposed that trait information is organized around two dimensions: leadership/competence and integrity/empathy. In a similar vein, Rahn, Aldrich, Borgida, and Sullivan (1990) proposed that voters differentiate between presidential task oriented traits and socioemotional oriented traits. Miller et al. (1986) proposed that traits are organized around five broad categories: compe­ tence, integrity, reliability, charisma, and personal characteristics (e.g., appear­ ance, background, health). Many researchers have assumed that every node composing this larger representation, both superordinate and specific, possesses an affective tag (Lodge, Stroh, & Wahlke, 1990; McGraw et al., 1991). The supposition that issue and trait information fall into separate clusters (McGraw et al., 1991) is not endorsed by all students of political cognition. In many cases, it is implicitly suggested that trait inferences are partially based on reactions to a candidate’s issue stands (Miller et al., 1986; Rahn et al., 1990). In addition, it should be recognized that trait organization typically is inferred on the basis of a factor analytic structure of specific trait judgments made about a candidate (Kinder, 1986; Rahn et al., 1990). Alternative methods that are more appropriate to ascertaining trait organization (e.g., recall, response time) may suggest an alternative structure. Ottati and Wyer (1992) and Wyer and Ottati (1992) applied the bin model of social memory to the domain of candidate representation. The model, which is metaphorical in nature, assumes that the information one receives for a particu­ lar purpose is stored in long-term memory in a content-addressable storage bin that refers to the person, object, or event to which the processing objective is relevant. Each bin is identified by a header. The header consists of a set of fea­ tures that are associated strongly with the referent and that, separately or in com­ bination, permit the bin to be identified. In the case of a political candidate, the header features may include the candidate’s name, political party, and personal appearance (i.e., visual image). To the extent that some features of the candidate elicit affective reactions, concepts of these reactions also might be included in the header. In certain respects, the bin header serves a function analogous to that of the superordinate person node in the hierarchical network models. Each cognitive representation that is formed from information about a refer­ ent constitutes a different knowledge configuration. Each knowledge configuration

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

171

(e.g., trait-behavior cluster) is stored in a bin on top of other configurations that have been deposited previously. The internal structure of these knowledge con­ figurations is organized in accordance with the dual-coding representational scheme outlined earlier. Although the information contained within a given knowledge configuration is associated in accordance with the network formulation, no such associative structure is assumed to exist between the knowledge configurations contained within the referent bin. One further consideration is important. A knowledge configuration is not stored in all of the bins to which it is relevant. Rather, it is deposited in only those bins whose referents are the focus of processing objectives that exist at the time it is formed or thought about. Thus, for example, suppose a representation is formed from a politician’s speech on disarmament. This representation might be stored in a politician bin if the listener’s goal is to form an impression of the politician, but in a disarmament bin if the listener’s goal is to learn or form an opinion about disarmament. In the latter case, the representation will not be retrieved if the recipient is asked later to evaluate the politician and searches for judgment-relevant information in the politician bin. Retrieving Candidate Information. As previously discussed, people who are required to make a judgment do not retrieve and use all of their knowledge about the judgment at hand. Instead, they sample only a subset of information. Politi­ cal cognition researchers have yet to present a rigorous model of the retrieval of candidate information. Wyer and Ottati (1992) drew on the Wyer and Srull bin model when delineat­ ing this retrieval process. They specified a priority system that governs the knowledge that people are likely to sample from memory. Specifically, assume that someone has formed several representations of a politician on the basis of information acquired at several different points in time and has stored these representations in a referent bin pertaining to the politician. If the recipient of this information later is asked to make a judgment or decision regarding the per­ son, the following steps are executed. The recipient first scans working memory for judgment-relevant information and, if this information is sufficient, uses it as a basis for judgment. If judgment-relevant information is not found in work­ ing memory, the recipient identifies the appropriate politician bin in long-term memory. If information contained in the bin header is sufficient, the recipient uses it as a basis for judgment without consulting the bin. Otherwise, the recipient performs a top-down search of the bin until a sufficient amount of judgment­ relevant information has been retrieved (see Wyer & Srull, 1986, for a more elaborate description). Integration and Candidate Evaluation. Political cognition researchers have introduced a variety of integration rules that attempt to model how citizens com­ bine information to arrive at an overall candidate evaluation. These models vary

172

IYENGAR AND OTTATI

along at least three dimensions: (a) the extent to which they emphasize category, exemplar, or piecemeal-based judgment; (b) the extent to which they propose that judgmental criteria are differentially versus equally weighted; and (c) the extent to which they propose that integration occurs on-line versus subsequent to retrieving candidate information. Piecemeal candidate evaluation integration rules presuppose that voters deliberatively consider a candidate’s specific, individuating attributes (e.g., issue posi­ tions) when arriving at an overall candidate evaluation. Algebraic models of piecemeal information integration within social psychology range from those that propose an averaging rule (Anderson, 1971, 1981) to those that propose an addi­ tive rule (Fishbein & Ajzen, 1975). Fishbein and Ajzen (1981) demonstrated that an additive rule strongly predicts candidate evaluation (attitude toward the candi­ date) as well as attitudes toward voting for a candidate. Additive models are also commonplace in the political literature. Kelley and Mirer (1974) proposed that after adding up their likes and dislikes associated with the candidates, voters choose the one with the highest net score. If a tie results, Kelley and Mirer proposed that citizens vote on the basis of their party identification. Similarly, Lodge et al. (1989) predicted candidate evaluation using the number of likes minus num­ ber of dislikes associated with the candidate. Other research has suggested that citizens sometimes are unable or unmoti­ vated to deliberatively consider a large set of piecemeal, individuating informa­ tion when arriving at a candidate evaluation. To the contrary, citizens often may evaluate a candidate on the basis of group stereotypes or category membership. For example, instead of evaluating a candidate on the basis of specific issue posi­ tions, voters may evaluate a candidate on the basis of the candidate’s party or racial group membership. Recent research has suggested that the tendency to en­ gage in piecemeal versus category-based candidate evaluation varies depending on the complexity of the judgment task and the amount of delay between infor­ mation exposure and judgment. When faced with a relatively simple, singular candidate judgment task involving little delay, individuals evaluate a candidate on the basis of a piecemeal consideration of specific issue stands (Ottati, 1990; Riggle, Ottati, Wyer, Kuklinski, & Schwarz, 1992). When faced with a more complex, comparative judgment task, category-based evaluation plays a signifi­ cant role (Ottati, 1990; Riggle et al., 1992). Delay between information exposure and judgment also can moderate the tendency to engage in category-based candi­ date evaluation (Riggle et al., 1992; Wyer et al., 1991). A second alternative to piecemeal candidate evaluation, the exemplar-based model, suggests that prior experience with a specific target person may deter­ mine judgments of a similar, subsequently encountered target politician. In this case, the candidate is not evaluated on the basis of category membership (e.g., partisanship), but on the basis of the voter’s evaluation of a similar previously encountered target person (Smith & Zarate, 1992). For example, Ronald Reagan may have been evaluated positively, “because he reminds me of John Wayne.”

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

173

Models that emphasize piecemeal candidate evaluation differ on the question of weighting. Kelly and Mirer (1983) explicitly proposed an equally weighted model. Others have suggested that each like or dislike should be weighed ac­ cording to the voter’s subjective certainty that the candidate possesses the attri­ bute (Fishbein & Ajzen, 1981), the subjective importance of the attribute (Krosnick, 1988; McGraw, Lodge, & Stroh, 1990), the cognitive accessibility of the attribute (Iyengar, Kinder, Peters, & Krosnick, 1984), or the serial posi­ tion in which the piece of information originally was encountered (Lodge & Stroh, 1992; McGraw et al., 1990). A third question concerns the time at which integration occurs. Survey research­ ers implicitly have assumed that candidate judgments are based on specific pieces of candidate information that are retrieved from long-term memory. That is, in­ tegration is assumed to occur after the retrieval of specific, judgment-relevant information. However, according to the on-line model of candidate evaluation, integration occurs during information exposure. That is, integration occurs be­ fore storage and representation in long-term memory. Specifically, the on-line model posits that the individual spontaneously culls the affective value from each message at the time of exposure, and then immediately integrates this affective value into a summary counter or tally that is stored in long-term memory. When a judgment is required, this on-line tally is retrieved from memory, rather than the specific pieces of information that contributed to it (Lodge et al., 1989; Lodge & Stroh, 1992; McGraw et al., 1990). Not all political research is necessarily consistent with the on-line model, however. For one thing, it is not clear how the on-line model handles issue prim­ ing effects that occur subsequent to exposure to candidate information. As noted previously, enhanced media coverage of an issue (even when not explicitly as­ sociated with a particular politician) can enhance the weight ascribed to that is­ sue when evaluating a candidate. Second, verbal protocols taken during exposure to candidate information rarely produce overall evaluative statements (Herstein, 1981). If voters continuously are updating their overall evaluation during exposure, they might be expected to mention overall evaluation with considerable frequen­ cy when producing these verbal protocols. Third, in some cases, retrieved can­ didate information is highly predictive of candidate evaluation. Specifically, when voters are cued to recall likes and dislikes associated with the candidate (evalua­ tive cued recall as opposed to free recall), the retrieved information strongly predicts candidate evaluation (McGraw et al., 1990). Herstein (1981) proposed a model that combines the implications of many of those introduced previously. If an individual formulates an overall negative can­ didate evaluation during information exposure, the voter decides against the nega­ tively evaluated candidate. If this does not occur, the voter compares the two candidates on a few salient attribute dimensions and chooses the candidate with the highest number of favorable comparisons. If this comparison results in a tie, the citizen votes on the basis of the party stereotype. This model is highly predictive

174

IYENGAR AND OTTATI

of vote choice (Herstein, 1981). There are, of course, a multitude of alternative possibilities. Payne, Bettman, and Johnson (1988) summarized a variety of decision rules that may be applicable to voter decisionmaking (e.g., dominance rule, lexico­ graphic rule, elimination by aspects). Political cognition researchers have yet to fully test these and other possibilities introduced by decisionmaking researchers. Affect and Political Judgment

Political judgments and decisions are influenced not only by descriptive political information, but also by the affective reactions one experiences in the course of processing this information en route to judgment. Here we focus on the role of episodic affective reactions, such as emotional responses to a candidate’s behavior and positive or negative mood states that are not necessarily attributed to any particular source. We intentionally exclude a consideration of evaluations and attitudes, which often are assumed to be relatively enduring or nonepisodic in nature. Affect and Political Information Processing. Episodic affect may influence political information processing in numerous ways. The first possibility is that episodic affective states influence concept accessibility. Given the central role of concept accessibility in political information processing, this sort of effect may shape processing en route to judgment significantly. A second possibility is that message recipients are not fully aware of the source of their affective reactions. Consequently, affective responses may be misattributed to the wrong political object, and thereby influence judgments of that object. Lastly, affective reactions may be stored in memory as a part of the individual’s representation of the mes­ sage referent. We discuss each of these possibilities in turn. The influence of affect on concept accessibility may play a role at two distinct stages of political information processing. The first may occur during exposure to a political communication. The second may occur when the individual later is asked to report a judgment. The initial possibility followed from the work of Bower, Gilligan, and Monteiro (1981). Their work suggested that concepts acti­ vated by subjects’ mood at the time information is presented leads them to en­ code selectively those aspects of the information to which the concepts were applicable. Specifically, positive mood subjects demonstrated a tendency to selec­ tively encode positive information, whereas negative mood subjects demonstrat­ ed a tendency to selectively encode negative information. Thus, one consequence of “feel good” campaign tactics may be to enhance the likelihood that voters selec­ tively encode positive information (e.g., agreeable issue statements) and ignore negative information (e.g., disagreeable issue statements). Although this possi­ bility merits further investigation, a preliminary study found no evidence of mood induced selective encoding effects for information presented about a political can­ didate (Ottati, 1992).

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

175

Affect also may influence concept accessibility at the time of judgment. For example, positive affective states may enhance the accessibility of positively evalu­ ated candidate information, whereas negative affective states may influence the accessibility of negatively evaluated candidate information. To the extent that judg­ ments of a candidate are based on accessible, specific pieces of information, this may bias the set of information on which the judgment is based. Although this second possibility has been supported by psychological evidence (Isen, Clark, Shalker, & Karp, 1978), preliminary research has found no evidence of this ef­ fect on candidate evaluation (Ottati, 1992). As noted previously, individuals may not be fully aware of the source of their affective responses. As a consequence, an affective reaction can be misattributed to an object that did not, in fact, actually elicit the affective response. Misattribution effects of this nature may occur either during exposure to a political mes­ sage or, alternatively, at the time of judgment. Feel good campaign techniques may capitalize on the misattribution of positive affect responses experienced during exposure to a candidate. For example, positive feelings elicited by patriotic mu­ sic or the American flag may be misattributed to a candidate or message. Although misattribution effects at the time of exposure have been demonstrated in other domains (Srull, 1983), they have yet to be investigated in the political domain. Misattribution effects also may occur at the time of judgment. Affective reac­ tions to some irrelevant event (e.g., the weather) may be misattributed to the object judged (e.g., a candidate) and operate as an additional source of informa­ tion when evaluating the political object. This sort of effect, which is not contin­ gent on biased retrieval, has been demonstrated for judgments of life satisfaction and marital satisfaction (Schwarz & Clore, 1983). Preliminary evidence has sug­ gested that it also may operate when evaluating a political candidate (Ottati & Wyer, 1992). Affective reactions, or concepts associated with them, also may be stored in memory as part of the individual’s representation of the message referent. For example, it is conceivable that emotional responses, or concepts associated with them (e.g., “He made me feel angry”), are stored in people’s memory as part of their representation of a political candidate. The notion that cognitive labels for emotional responses may function as attributes in a perceiver’s cognitive representation of a target person already has been introduced by cognitive the­ orists (Bower, 1981; Wyer & Srull, 1989). Episodic emotional reactions that are associated strongly with a candidate may be associated with the superordinate person node or header of the candidate representation (Ottati & Wyer, 1991; Wyer & Srull, 1989). Emotional response labels that are less central to the candidate representation may be similar in status to trait nodes, which are associated with subordinate behaviors that elicited the emotion. Alternatively, emotional response labels may occupy a status similar to episodic behaviors, which are organized around a su­ perordinate trait node in a trait-behavior knowledge configuration. These latter

176

IYENGAR AND OTTATI

two possibilities have not been distinguished clearly by cognitive theorists, much less students of political cognition. In any case, it seems likely that emotional responses sometimes are represented in memory, even though specific aspects of the communication that initially gave rise to the reactions are not retained. Thus, for example, people may recall that a particular candidate’s speech made them angry without being able to remember what specific things the speaker said that gave rise to this reaction. Cognition and Affect: Independent Determinants o f Political Judgment ? Cur­ rent research has suggested that beliefs and emotions serve as unique determinants of political attitudes (Abelson, Kinder, Peters, & Fiske, 1982; Conover & Feld­ man, 1986). This finding emerged in a variety of political domains; including candidate evaluation (Abelson et al., 1982), evaluations of economic conditions (Conover & Feldman, 1986), and expressions of political tolerance (Kuklinski et al., 1991; Ottati et al., 1989). This evidence is consistent with Rosenberg and Hovland’s (1960) three-component view of attitudes as consisting of beliefs, feel­ ings, and behavioral response tendencies. Typically, it is assumed that these three components are not always highly related. Consistent with this perspective, many psychologists have obtained evidence for the discriminant validity of scales meas­ uring the three attitudinal components (Breckler & Wiggins, 1989; Kothandapani, 1971; Millar & Tesser, 1989). Recent political research demonstrated that emotional responses operate as predictors of attitudes independently of beliefs (Abelson et al., 1982; Conover & Feldman, 1986; Ottati, Steenbergen, & Riggle, 1992). This research also demonstrated that emotions and beliefs exhibit differences in internal structure. Although positive and negative beliefs about an object are related inversely, posi­ tive and negative emotions are relatively independent (Abelson et al., 1982; Conover & Feldman, 1986). Taken together, these findings suggested that be­ liefs and emotions operate as partially distinct determinants of political attitudes. There are several psychological process mechanisms that plausibly could un­ derlie these findings. One rests on the assumption that beliefs determine an in­ dividual’s emotional responses at the time that they are experienced, but cannot be retrieved when the individual later is asked to respond to the questionnaire. For example, an individual who feels uneasy about a politician’s remark may forget quickly that it was the perception that the remark was dishonest that originally triggered the feeling. Such an individual may endorse the uneasy emotion item, but fail to endorse the dishonest belief item. As a consequence, this particular emotion may contribute unique information when predicting candidate evaluation. This explanation is both intuitively plausible and indirectly supported by related research. Individuals often recall feeling an emotional response without being able to recall the specific beliefs or cognitions that originally gave rise to the feel­ ing. Related to this argument, current research indicates that individuals often make on-line judgments when encoding information about political actors (Hastie

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

177

& Park, 1986; Lichtenstein & Srull, 1987; Lodge et al., 1989). Similarly, Petty and Cacioppo (1986) argued that perceivers often generate cognitive responses when receiving information about an attitude object. These on-line judgments may be transferred to long-term memory independently of the specific perceptual be­ liefs that triggered them. The processing of emotional responses may operate in an analogous fashion. That is, an emotional response to a candidate may be stored in memory independent of the specific perceptual belief that originally contribut­ ed to the emotional response. Later, the individual may recall the emotional response without being able to retrieve the specific belief that triggered the emo­ tional response. As a consequence, this emotional response may contain evaluation­ relevant information that is not accounted for by any beliefs that are accessible to the respondent. A second line of reasoning rests on the assumption that conscious beliefs do not necessarily determine emotional responses when they are experienced origi­ nally. This may occur because emotional responses arise without any prior cog­ nitive appraisal whatsoever (Wilson, 1979; Zajonc, 1984), or because cognitive appraisals occur automatically or below the level of conscious awareness (Bargh, 1985; Ottati & Wyer, 1992). Either of these possibilities may lead emotional responses to contain evaluation-relevant information that is not accounted for by beliefs that are accessible to the respondent. However, the claim that emotional responses operate as autonomous predic­ tors of political attitudes has been challenged. Ottati, Steenbergen, and Riggle (1992) noted that the evidence suggesting that positive and negative emotions are more independent than positive and negative beliefs only arises when employing closed-ended dichotomous measures of emotion. When employing alternative measures, this distinction between beliefs and emotions fails to emerge (Ottati, 1992; Ottati, Steenbergen, & Riggle, 1992). Also, freely generated cognitive ex­ planations for emotional responses fully account for the role of emotions when predicting attitudes toward a candidate (Ottati & Wyer, 1992). Of course, these cognitive explanations simply may reflect justifications for emotional responses that were triggered by other criteria (Nisbett & Wilson, 1977). Nevertheless, these recent findings suggest that the debate regarding the uniqueness of cognitive and affective determinants of political attitudes is yet to be fully resolved.

FUTURE DIRECTIONS

It is clear that much of the work we have covered in this chapter represents a significant advance in the understanding of political cognition. However, several remaining concerns need to be explored further. In this concluding section, we identify problem areas that are particularly deserving of attention from cogni­ tively inclined political psychologists. In most social cognition experiments, the researcher controls the amount and

178

IYENGAR AND OTTATI

nature of the information presented to the individual. That is, exposure is largely involuntary; the individual is not permitted to seek out freely or avoid exposure to certain types of information. Yet the real world of politics offers the citizen countless opportunities to actively seek out or avoid certain political messages. Citizens exercise this voluntary control over information exposure whenever they read certain news articles while ignoring others, when they switch television chan­ nels, and so on. Although many political behavior theorists have assumed that citizens seek out exposure to information that is consistent with prior attitudes and expecta­ tions, there is virtually no empirical documentation of this tendency. Process trac­ ing techniques geared toward mapping the information seeking process may be particularly useful in this regard. Herstein (1981) analyzed verbal protocol data in an attempt to map this process. Other methodological techniques, such as presenting the individual with an “information board” that enables the individual to select the order and amount of information used for judgment, may offer fur­ ther insights into this process. Once a political message has been received and comprehended, it is encoded in terms of relevant, accessible concepts. Political cognition researchers already have noted that prior knowledge about a candidate (e.g., partisanship) influences concept accessibility, which can produce selective encoding effects. The role of individual differences in chronic accessibility (e.g., partisan schematics vs. aschematics) also has been documented empirically. Less attention has been given to the role of situational priming occurring immediately prior to information ex­ posure. Basic social cognition research has suggested that priming a concept can influence the interpretation of subsequently presented ambiguous information. Thus, the interpretation of a televised news story (e.g., bombing Libya) may be influenced by the nature of the preceding story (e.g., attacking drug lords vs. Hitler’s bombing of England). This sort of priming effect, which occurs at a rela­ tively early stage of processing, has yet to be documented in the political arena. Perhaps the most neglected aspect of political cognition research centers around the question of how political information is represented in long-term memory. As noted previously, few have attempted to present an explicit model of the ar­ chitecture of issue- or policy-based considerations (Judd & Krosnick, 1989; McGraw & Pinney, 1990). Attempts to model the linkages between specific is­ sues and superordinate core values have been based largely on judgment data. Here the strength of the association between a crowning posture (e.g., individu­ alism or self-reliance) and a specific issue (e.g., affirmative action) often is oper­ ationalized in terms of the correlation between individual endorsements of the specific issue and core value. Needless to say, correlational data do not provide an ideal basis for inferring the representational structure of social information. An analysis of open-ended verbal protocols, free recall, or response time data potentially could be applied to ascertain the organization and structure of issue information represented in long-term memory. Basic research has suggested that

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

179

such techniques often reveal aspects of memory structure that simply are not reflected in judgment data. As such, an explicit model of the issue belief system as a cognitive structure has yet to be articulated. An analogous state of affairs exists when considering past research concern­ ing the representation of political actors in memory. A considerable amount of work has sought to ascertain the factor structure of emotional responses and trait ascriptions to a candidate. Again, associational structure is operationalized in terms of the correlational structure of judgments about a political actor. Only recently have researchers analyzed recall data with the specific objective of ascertaining the representational structure of candidate information (McGraw et al., 1991). Further applications of such cognitive indicators (e.g., response time) are needed. Models of information integration and inference have been explored more fully by students of political cognition. Many recent advances in this area have been psychologically realistic, because they have kept the voter’s cognitive limits in mind. Examples include the recent emphasis on category-based and on-line can­ didate evaluation, as well as piecemeal integration models that acknowledge that judgments are based on a limited, accessible set of information. Political cogni­ tion research has suffered no shortage of information integration models. A question that is developed less fully concerns the conditions that moderate the tendency for voters to employ one or another route to judgment. For exam­ ple, the tendency to engage in on-line processing is contingent on voter political sophistication. Also, the tendency to engage in category-based candidate evalua­ tion appears to depend on the complexity of the judgment task. Yet, it seems probable that additional contingencies exist. Preliminary evidence suggests that the tendency to engage in category-based candidate evaluation is moderated by the quality of nonverbal cues that accompany or precede a candidate’s verbal mes­ sage (Ottati, Steenbergen, & Riggle, 1992; Wyer et al., 1991). It is also possible that such effects are contingent on the modality (e.g., print vs. radio) or manner of framing the presented information. These and other contingencies need to be explored more fully. Many of the previous suggestions involve applying more cognitively oriented methods to the study of political cognition. In a number of cases, political cogni­ tion researchers have adopted cognitive terminology without capitalizing on the methodological advances characteristic of the information processing approach. To be a truly interdisciplinary field, political psychology needs to incorporate both the methodological and theoretical advances associated with the cognitive revolution. We encourage further research that advances us to this end. REFERENCES Abelson, R. P., Kinder, D. R., Peters, M. D., & Fiske, S. T. (1982). Affective and semantic com­ ponents in political person perception. Journal o f Personality and Social Psychology, 42, 619-630. Adorno, T. E., Frenkel-Brunswik, E., Levinson, D. & Sanford, R. N. (1950). The authoritarian personality. NY: Harper & Bros.

180

IYENGAR AND OTTATI

Anderson, J. R. (1983). The architecture o f cognition. Cambridge: Harvard University Press. Anderson, N. H. (1965). Average versus adding as a stimulus-combination rule in impression for­ mation. Journal o f Experimental Psychology, 70, 394-400. Anderson, N. H. (1981). Foundations o f information integration theory. NY: Academic Press. Anderson, N. H. (1971). Integration theory and attitude change. Psychological Review, 78, 171-206. Anderson, R. C., Reynolds, R. E., Schallert, D. L., & Goetz, E. T. (1976). Frameworks fo r com­ prehending discourse (Tech. Rep. No. 12). Urbana, IL: Laboratory of Cognitive Studies in Edu­ cation, University of Illinois at Urbana-Champaign. Ansolabehere, S. (1988, April). Rational choice and the puzzle o f negative voting: A random utilities model o f the vote. Paper presented at the Midwest Political Science Association meeting, Chicago. Ansolabehere, S., Iyengar, S., & Simon, A. (1990, August). G ood news, bad news, and economic voting. Paper presented at the annual meeting of the American Political Science Association, San Francisco. Barber, J. D. (1985). Presidential character: Predicting performance in the White House. EnglewoodCliffs, NJ: Prentice-Hall. Bargh, J. A. (1985). Automatic and conscious processing of social information. In R. E. Wyer, Jr. & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 3, pp. 1-43). Hillsdale, NJ: Lawrence Erlbaum Associates. Bargh, J. A., Bond, R., Lombardi, W ., & Tota, M. E. (1986). The additive nature of chronic and temporary sources of construct accessibility. Journal o f Personality and Social Psychology, 50, 869-878. Bartels, L. M. (1985). Expectations and preferences in presidential nominating campaigns. Ameri­ can Political Science Review, 79, 804-815. Bartels, L. M. (1988). Presidential primaries and the dynamics o f public choice. Princeton, NJ: Prince­ ton University Press. Behr, R. L., & Iyengar, S. (1985). Television news, real-world cues, and changes in the public agenda. Public Opinion Quarterly, 49, 38-57. Berelson, B., Lazarsfeld, P. F., & McPhee, W. N. (1954). Voting: A study o f opinion formation in a presidential campaign. Chicago: University of Chicago Press. Bishop, G. F., Oldendick, R. W ., & Tuchfarber, A. J. (1982). Political information processing: Question order and context effects. Political Behavior, 4, 177-200. Bloom, H. S., & Price, H. D. (1975). Voter response to short-run economic conditions. American Political Science Review, 69, 1240-1253. Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129-148. Bower, G. H., Gilligan, S. G., & Monteiro, K. P. (1981). Selectivity of learning caused by affective states. Journal o f Experimental Psychology: General, 110, 451-483. Brady, H. E., & Ansolabehere, S. (1989). The nature of utility functions in mass publics. American P olitical Science Review, 83, 143-163. Breckler, S. J., & Wiggins, E. (1989). Affect versus evaluation in the structure of attitudes. Journal o f Experimental Social Psychology, 25, 253-271. Brody, R. A., & Page, B. I. (1972). The assessment of policy voting. American Political Science Review, 66, 450-458. Bruner, J. S. (1951). Personality dynamics and the process of perceiving. In R. Blake & G. Ramsey (Eds.), Perception: An approach to personality. New York: Ronald Press. Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American voter. New York: Wiley. Campbell, A., Gurin, G., & Miller, W. E. (1954). The voter decides. Evanston, IL: Row & Peterson. Carver, C. C., Ganellen, R., Froming, W. J., & Chambers, W. (1983). Modelling: An analysis in terms of category accessibility. Journal o f Experimental Social Psychology, 19, 403-421. Chappell, H. W ., & Keech, W. (1990). Citizen information, rationality, and the politics of macroeconomic policy. In J. Ferejohn & J. Kuklinski (Eds.), Information and dem ocratic processes (pp. 313-328). Urbana, IL: University of Illinois Press.

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

181

Collins, A. M ., & Loftus, E. F. (1975). A spreading activation theory of semantic processing. P sy­ chological Review, 82, 407-428. Collins, A. M ., & Quillian, M. R. (1969). Retrieval time from semantic memory. Journal o f Verbal Learning and Verbal Behavior, 8, 240-247. Conover, P., & Feldman, S. (1984). How people organize the political world: A schematic model. American Journal o f Political Science, 28, 95-126. Conover, P., & Feldman, S. (1986). Emotional responses to the economy: I’m mad as hell and I’m not going to take it anymore. American Journal o f Political Science, 30, 50-78. Converse, P. E. (1964). The nature of belief systems in mass publics. In D. Apter (Ed.), Ideology and discontent (pp. 206-261). New York: Free Press. Downs, A. (1957). An economic theory o f dem ocracy. New York: Harper & Row. Enelow, J., & Hinich, M. (1984). Spatial models o f voting. New York: Cambridge University Press. Erbring, L., Goldenberg, E., & Miller, A. (1980). Front-page news and real-world cues: A new look at agenda-setting. American Journal o f Political Science, 24, 16-49. Fazio, R. H. (1990). Multiple processes by which attitudes guide behavior: The MODE model as an integrative framework. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 75-109). New York: Academic Press. Fazio, R. H., & Williams, C. J. (1986). Attitude accessibility as a moderator of the attitude-perception and attitude-behavior relations: An investigation of the 1984 presidential election. Journal o f Per­ sonality and Social Psychology, 51, 505-514. Fiorina, M. P. (1981). Retrospective voting in American national elections. New Haven, CT: Yale University Press. Fiorina, M. P., & Shepsle, K. A. (1989). Is negative voting an artifact? American Journal o f P oliti­ cal Science, 33, 423-439. Fishbein, M ., & Ajzen, I. (1975). Belief, attitude, intention, and behavior. Manila, Philippines: Addison-Wesley. Fishbein, M ., & Ajzen, I. (1981). Attitudes and voting behavior: An application of the theory of reasoned action. In G. M. Stephenson & J. M. Davis (Eds.), Progress in applied social psycholo­ gy (pp. 253-313). Chichester, England: Wiley. Fiske, S. T ., & Kinder, D. R. (1981). Involvement, expertise, and schema use: Evidence from polit­ ical cognition. In N. Cantor & J. Kihlstrom (Eds.), Personality, cognition, and social interaction (pp. 171-190). Hillsdale, NJ: Lawrence Erlbaum Associates. Fiske, S. T ., Kinder, D. R., & Larter, W. (1983). The novice and the expert: Knowledge based strategies in political cognition. Journal o f Experimental Social Psychology, 19, 381-400. Fiske, S. T., & Pavelchak, M. A. (1986). Category-based vs. piecemeal-based affective responses: Developments in schema-triggered affect. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook o f motivation and cognition (pp. 167-203). New York: Guilford. George, A. L., & George, J. L. (1964). Woodrow Wilson and Colonel House: A personality study. New York: Dover. Graber, D. (1988). Processing the news: H ow people tame the information tide. NY: Longman. Grether, D., & Plott, C. (1979). Economic theory of choice and the preference reversal phenome­ non. American Economic Review, 69, 623-638. Hamilton, D. L ., Driscoll, D., & Worth, L. T. (1989). Cognitive organization of impressions: Ef­ fects of incongruency in complex representations. Journal o f Personality and Social Psychology, 57, 925-939. Hastie, R. (1980). Memory for behavioral information that confirms or contradicts a personality im­ pression. In R. Hastie, T. M. Ostrom, E. B. Ebbesen, R. S. Wyer, Jr., D. L. Hamilton, & D. E. Carlston (Eds.), Person memory: The cognitive basis o f social perception (pp. 155-177). Hillsdale, NJ: Lawrence Erlbaum Associates. Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the judgment task is memory-based or on-line. Psychological Review, 93, 258-268.

182

IYENGAR AND OTTATI

Herr, P. M ., Sherman, S., & Fazio, R. (1983). On the consequences of priming: Assimilation and contrast effects. Journal o f Experimental Social Psychology, 19, 323-340. Herstein, J. A. (1981). Keeping the voter’s limits in mind: A cognitive process analysis of decision making in voting. Journal o f Personality and Social Psychology, 40, 843-861. Hibbs, D. A. (1987). The American political economy. Cambridge, MA: Harvard University Press. Higgins, E. T ., & Chaires, W. (1980). Accessibility of interrelational constructs: Implications for stimulus encoding and creativity. Journal o f Experimental Social Psychology, 16, 348-361. Higgins, E. T., & King, G. (1981). Category accessibility and information-processing: Consequences of individual and contextual variability. In N. Cantor & J. Kihlstrom (Eds.), Personality, cogni­ tion and social interaction (pp. 69-122). Hillsdale, NJ: Lawrence Erlbaum Associates. Hurwitz, J., & Peffley, M. A. (1987). How are foreign policy attitudes structured? A hierarchical model. American Political Science Review, 81, 1099-1120. Isen, A. M., Clark, M ., Shalker, T. E., & Karp, L. (1978). Affect, accessibility of material in memory, and behavior: A cognitive loop? Journal o f Personality and Social Psychology, 36, 1-12. Iyengar, S. (1990a). Shortcuts to political knowledge: Selective attention and the accessibility bias. In J. Ferejohn & J. Kuklinski (Eds.), Information and dem ocratic processes (pp. 160-185). Urbana, IL: University of Illinois Press. Iyengar, S. (1990b). The accessibility bias in politics: Television news and public opinion. Interna­ tional Journal o f Public Opinion Research, 2, 1-15. Iyengar, S. (1991). Is anyone responsible? H ow television fram es political issues. Chicago: Univer­ sity of Chicago Press. Iyengar, S., & Kinder, D. R. (1986). More than meets the eye: Television news, priming, and presiden­ tial evaluations. In G. Comstrock (Ed.), Public communication and behavior (Vol. 1, pp. 135-171). New York: Academic Press. Iyengar, S., & Kinder, D. R. (1987). News that matters: Television and American opinion. Chicago: University of Chicago Press. Iyengar, S., Kinder, D. R., Peters, M. D., & Krosnick, J. A. (1984). The evening news and presidential evaluations. Journal o f Personality and Social Psychology, 46, 778-787. Iyengar, S., & McGuire, W. J. (Eds.). (1993). Explorations in political psychology. Durham, NC: Duke University Press. Iyengar, S., & Simon, A. (1992, August). News coverage o f the Gulf crisis and public opinion: A survey o f effects. Paper presented at the annual meeting of the International Communication As­ sociation, Miami. Judd, C. M ., & Krosnick, J. A. (1989). The structural bases of consistency among political atti­ tudes: Effects of political expertise and attitude importance. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 99-128). Hillsdale, NJ: Lawrence Erlbaum Associates. Kahneman, D., Slovic, P., & Tversky A. (Eds.). (1982). Judgment under uncertainty: Heuristics and biases. New York: Cambridge University Press. Kahneman, D., & Tversky, A. (1982). The psychology of preferences. Science, 246, 136-142. Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. American Psychologist, 39, 341-350. Kahneman, D., & Tversky, A. (1987). Rational choice and the framing of decisions. In M. W. Redes & R. Hogarth (Eds.), Rational choice: The contrast between economics and psychology (pp. 67-94). Chicago: University of Chicago Press. Kelley, S., & Mirer, T. (1974). The simple act of voting. American Political Science Review, 61, 572-591. Kemell, S. (1977). Presidential popularity and negative voting: An alternative explanation of the midterm congressional decline of the President’s party. American Political Science Review, 71, 44-66. Key, V. O., Jr. (1966). The responsible electorate. Cambridge, MA: Harvard University Press. Kiewiet, R. D. (1983). M acroeconomic and micropolitics. Chicago: University of Chicago'Press.

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

183

Kiewiet, R. D., & Rivers, R. D. (1984). A retrospective on retrospective voting. Political Behavior, 6, 369-393. Kinder, D. R. (1981). Presidents, prosperity, and public opinion. Public Opinion Quarterly, 45, 1-21. Kinder, D. R. (1986). Presidential character revisited. In R. R. Lau & D. O. Sears (Eds.), Political cognition: The 19th annual Carnegie symposium on cognition (pp. 233-255). Hillsdale, NJ: Lawrence Erlbaum Associates. Kinder, D. R., & Kiewiet, R. D. (1979). Economic discontent and political behavior: The role of personal grievances and collective judgments in congressional voting. American Journal o f P oliti­ cal Science, 23, 495-527. Kinder, D. R., & Mebane, W. R., Jr. (1983). Politics and economics in everyday life. In K. Monroe (Ed.), The political process and economic change. New York: Agathon. Klinger, E. (1977). Meaning and void: Inner experience and the incentives in p eo p le’s lives. Min­ neapolis: University of Minnesota Press. Kothandapani, V. (1971). Validation of feeling, belief, and intention to act as three components of attitude and their contribution to prediction of contraceptive behavior. Journal o f Personality and Social Psychology, 19, 321-333. Krosnick, J. A. (1988). The role of attitude importance in social evaluation: A study of policy prefer­ ences, presidential candidate evaluations, and voting behavior. Journal o f Personality and Social Psychology, 55, 196-201. Krosnick, J. A. (1989). Attitude importance and attitude accessibility. Personality and Social P sy­ chology Bulletin, 15, 297-308. Krosnick, J. A., & Kinder, D. R. (1990). Altering the foundations of popular support for the presi­ dent through priming. American Political Science Review, 84, 497-512. Kuklinski, J. H., Riggle, E., Ottati, V., Schwarz, N., & Wyer, R. S. (1991). The basis of political tolerance judgments: Cognition or emotion? American Journal o f Political Science, 35, 1-27. Lachman, R., Lachman, J. T ., & Butterfield, E. C. (1979). Cognitive psychology and information processing: An introduction. Hillsdale, NJ: Lawrence Erlbaum Associates. Lane, R. E. (1962). Political ideology: Why the American common man believes what he does. New York: Free Press. Lang, K., & Lang, G. (1959). The mass media and voting. In E. J. Burdick & A. J. Brodbeck (Eds.), American voting behavior. New York: Free Press. Lasswell, H. D. (1948). Pow er and personality. New York: Norton. Lau, R. L. (1982). Negativity in political perception. Political Behavior, 4, 353-378. Lau, R. L. (1985). Two explanations for negativity effects in political behavior. American Journal o f Political Science, 29, 119-138. Lau, R., & Erber, R. (1985). Political sophistication: An information processing perspective. In S. Kraus & R. M. Perloff (Eds.), Mass media and political thought: An information processing ap­ proach (pp. 37-64). Beverly Hills, CA: Sage. Lau, R. R., Smith, R. A., & Fiske, S. T. (1991). Political beliefs, policy interpretations, and politi­ cal persuasion. The Journal o f Politics, 58, 644-675. Lazarsfeld, P., Berelson, B., & Gaudet, H. (1949). The people’s choice (2nd ed.). New York: Columbia University Press. Lewis-Beck, M. (1988). Economics and election: The major western democracies. Ann Arbor: Univer­ sity of Michigan Press. Lichtenstein, M ., & Srull, T. K. (1987). Processing objectives as a determinant of the relationship between recall and judgment. Journal o f Experimental Social Psychology, 23, 93-118. Lingle, J. H., Geva, N., Ostrom, T. M ., Leippe, M. R., & Baumgardner, M. H. (1979). Thematic effects of person judgments on impression organization. Journal o f Personality and Social P sy­ chology, 37, 674-687. Lodge, M ., & Hamill, R. (1986). A partisan schema for political information processing. American Political Science Review, 80, 505-519.

184

IYENGAR AND OTTATI

Lodge, M ., McGraw, K., & Stroh, P. (1989). An impression-driven model of candidate evaluation. American Political Science R eview, 83, 399-419. Lodge, M ., & Stroh, P. (1993). Inside the mental voting booth: An impression-driven model of can­ didate evaluation. In S. Iyengar & W. J. McGuire (Eds.), Explorations in political psychology (pp. 225-263). Durham, NC: Duke University Press. Lodge, M ., Stroh, P., & Wahlke, J. (1990). Black-box models of evaluation. Political Behavior, 12, 5-18. MacKuen, M. B. (1981). Social communication and the mass policy agenda. In M. MacKuen & S. Coombs (Eds.), More than news: Media pow er in pu blic affairs (pp. 19-144). Beverly Hills, CA: Sage. MacKuen, M. B. (1983). Political drama, economic conditions, and the dynamics of presidential popularity. American Journal o f Political Science, 27, 165-192. Manis, M ., Nelson, T ., & Shedler, J. (1988). Stereotypes and social judgment: Extremity, assimila­ tion, and contrast. Journal o f Personality and Social Psychology, 55, 28-36. Marcus, G. E ., Tabb, D., & Sullivan, J. L. (1974). The application of individual differences scaling to the measurement of political ideologies. American Journal o f Political Science, 18, 405-420. Markus, G. B. (1988). The impact of personal and national economic conditions on the presidential vote: A pooled cross-sectional analysis. American Journal o f Political Science, 32, 137-154. Martin, L. L., Seta, J. J., & Crelia, R. A. (1990). Assimilation and contrast as a function of people’s willingness and ability to expend effort in forming an impression. Journal o f Personality and So­ cial Psychology, 59, 27-37. McClelland, D. C. (1961). The achieving society. Princeton, NJ: Van Nostrand. McGraw, K., Lodge, M ., & Stroh, P. (1990). Order effects in the evaluation of political candidates. Political Behavior, 12, 41-58. McGraw, K., & Pinney, N. (1990). The effects of general and domain-specific expertise on political memory and judgment. Social Cognition, 8, 9-30. McGraw, K., Pinney, N., & Neumann, D. (1991). Memory for political actors: Contrasting the use of semantic and evaluative organizational strategies. Political Behavior, 13, 165-189. McGuire, W. J. (1992). The poli.-psy. relationship: Three phases of a long affair. In S. Iyengar & W. J. M cGuire^Eds.), Explorations in political psychology (pp. 9-35). Durham, NC: Duke University Press. McGuire, W. J., & McGuire, C. V. (1991). The content, structure and operation of thought sys­ tems. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Advances in social cognition (Vol. 4, pp. 1-78). Hillsdale, NJ: Lawrence Erlbaum Associates. McNeil, B., Parker, S., Sox, H., Jr., & Tversky, A. (1982). On the elicitation of preferences for alternative therapies. New England Journal o f Medicine, 306, 1259-1262. Millar, M. G., & Tesser, A. (1989). The effects of affective-cognitive consistency and thought on the attitude-behavior relation. Journal o f Experimental Social Psychology, 25, 189-202. Miller, A. H., Wattenberg, M. P., & Malanchuk, O. (1986). Schematic assessments of presidential candidates. American Political Science Review, 80, 521-540. Nie, N. H., & Anderson, K. (1974). Mass belief systems revisited: Political change and attitude structure. Journal o f Politics, 36, 540-591. Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings o f social ju dg­ ment. Englewood Cliffs, NJ: Prentice-Hall. Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we know: Verbal reports on mental process­ es. Psychological Review, 84, 231-259. Norpoth, H., Lewis-Beck, M ., and Lafay, J. (Eds.). (1991). Economics and politics: The calculus o f support. Ann Arbor: University of Michigan Press. Ostrom, C. W ., & Simon, D. M. (1989). The man in the Teflon suit? Public Opinion Quarterly, 53, 353-387. Ottati, V. (1990). Determinants of political judgments: The joint influence of normative and heuris­ tic rules of inference. Political Behavior, 12, 159-179.

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

185

Ottati, V., Fishbein, M ., & Middlestadt, S. E. (1988). Determinants of voters’ beliefs about the can­ didate’s stands on the issues: The role of evaluative bias heuristics and the candidates’ expressed message. Journal o f Personality and Social Psychology, 57, 517-529. Ottati, V., Riggle, E ., Wyer, R. S., Jr., Schwarz, N., & Kuklinski, J. (1989). The cognitive and affective bases of opinion survey responses. Journal o f Personality and Social Psychology, 545, 404-415. Ottati, V., Steenbergen, M ., & Riggle, E. (1992). The cognitive and affective components of politi­ cal attitudes: Measuring the determinants of candidate evaluations. Political Behavior, 14, 423-442. Ottati, V ., Terkildsen, N., & Hubbard, C. (1992). Cognitive processes underlying beliefs about a candidate’s stands on the issues. Unpublished manuscript, State University of New York, Stony Brook. Ottati, V., & Wyer, R. S., Jr. (1992). Affect and political judgment. In S. Iyengar & W. J. McGuire (Eds.), Explorations in political psychology (pp. 296-320). Durham, NC: Duke University Press. Page, B. I. (1976). A theory of political ambiguity. American Political Science Review, 70, 742-752. Payne, J. W ., Bettman, J., & Johnson, E. J. (1988). Adaptive strategy selection in decision making. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 14, 534-552. Payne, J. W ., Laughunn, D., & Crum, R. (1980). Translation of gambles and aspiration level ef­ fects in risky choice behavior. Management Science, 26, 1039-1060. Peffley, M ., & Hurwitz, J. (1985). A hierarchical model of attitude constraint. American Journal o f Political Science, 29, 871-890. Peffley, M ., & Williams, J. T. (1985). Attributing presidential responsibility for national economic problems. American Politics Quarterly, 13, 393-425. Pennington, N ., & Hastie, R. (1986). Evidence evaluation in complex decision making. Journal o f Personality and Social Psychology, 51, 242-258. Petrocik, J. R. (1990). The theory o f issue ownership: Issues, agendas, and electoral coalitions in the 1988 election. Unpublished manuscript, University of California, Los Angeles. Petty, R. E ., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlag. Petty, R. E. Ostrom, T. M ., & Brock, T. C. (1981). Cognitive responses in persuasion. Hillsdale, NJ: Lawrence Erlbaum Associates. Quattrone, G. A., & Tversky, A. (1988). Contrasting rational and psychological analyses of politi­ cal choice. American Political Science Review, 82 , 719-736. Radcliff, B. (1986, April). Solving a puzzle: Aggregate analysis and economic voting revisited. Paper presented at the annual meeting of the Midwestern Political Science Association, Chicago. Rahn, W. M ., Aldrich, J. H., Borgida, E., & Sullivan, J. L. (1990). A social-cognitive model of candidate appraisal. In J. Ferejohn & J. Kuklinski (Eds.), Information and dem ocratic processes (pp. 136-159). Urbana, IL: University of Illinois Press. Riggle, E., Ottati, V., Kuklinski, J., Wyer, R. S., & Schwarz, N. (1989). Bases o f political ju d g ­ ment: The role o f a candidate’s physical attractiveness, party membership, and voting record. Un­ published manuscript, University of Illinois, Champaign. Riker, W. J. (1986). The art o f political manipulation. New Haven, CT: Yale University Press. Rivers, R. D. (n.d). M icroeconomics and m acropolitics: A solution to the Kram er problem . Unpub­ lished manuscript, Stanford University, Stanford, CA. Robinson, M ., & Sheehan, M. A. (1980). Over the wire and on television. New York: Basic Books. Roenker, D. L., Thompson, C. P., & Brown, S. C. (1971). Comparison of measures for the estima­ tion of free recall. Psychological Bulletin, 76, 45-48. Rogers, E. M ., & Dearing, J. W. (1988). Agenda-setting research: Where has it been and where is it going? In J. A. Anderson (Ed.), Communication yearbook (Vol. 11, pp. 555-594). Beverly Hills, CA: Sage. Rosenberg, M. J., & Hovland, C. I. (1960). Cognitive, affective, and behavioral components of atti­ tudes. In M. J. Rosenberg & C. I. Hovland (Eds.), Attitude organization and change (pp. 1-14). New Haven, CT: Yale University Press.

186

IYENGAR AND OTTATI

Schlozman, K., & Verba, S. (1979). Insult to injury: Unemployment, class, and political response. Cambridge: Harvard University Press. Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: Informa­ tive and directive functions of affective states. Journal o f Personality and Social P sychology, 45, 513-523. Sears, D. O., & Funk, C. L. (1991). The role of self-interest in social and political attitudes. In M. Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 24, pp. 1-91). Orlando, FL: Academic Press. Shepsle, K. A. (1972). The strategy of ambiguity: Uncertainty and electoral competition. American Political Science Review, 66, 555-568. Slovic, P., Fischhoff, B., & Lichtenstein, S. (1982). Response mode, framing, and informationprocessing effects in risk assessment. In R. Hogarth (Ed.), New directions f o r methodology o f so­ cial and behavioral science: Question fram ing and response consistency (pp. 21-36). San Francis­ co: Jossey-Bass. Smith, E. R. (1990). Content and process specificity in the effects of prior experiences. In T. K. Srull & R. S. Wyer, Jr. (Eds.), Advances in social cognition (Vol. 3, pp. 1-59). Hillsdale, NJ: Lawrence Erlbaum Associates. Smith, E. R., & Zarate, M. A. (1992). Exemplar-based models of social judgment. Psychological Review, 99, 3-21. Smith, T. (1987). That which we call welfare by any other name would smell sweeter: An analysis of the impact of question wording on response patterns. Public Opinion Quarterly, 51, 75-83. Sniderman, P. M ., & Brody, R. A. (1978). Coping: The ethic of self-reliance. American Journal o f Political Science, 21, 501-521. Snyder, M ., & Swann, W. B. (1978). Hypothesis-testing processes in social interaction. Journal o f Personality and Social Psychology, 36, 1202-1212. Srull, T. K. (1981). Person memory: Some tests of associative storage and retrieval models. Journal o f Experimental Psychology: Human Learning and Memory, 7, 440-462. Srull, T. K. (1983). Affect and memory: The impact of affective reactions in advertising on the represen­ tation of product information in memory. Advances in Consumer Research, 10, 572-576. Srull, T. K., & Wyer, R. S., Jr. (1979). The role of category accessibility in the interpretation of information about persons: Some determinants and implications. Journal o f Personality and So­ cial Psychology, 37, 1660-1672. Srull, T. K., & Wyer, R. S., Jr. (1989). Person memory and judgment. Psychological Review, 96, 58-83. Sullivan, J. L., Pierson, J. E., & Marcus, G. E. (1978). Ideological constraint in the mass public: A methodological critique and some new findings. American Journal o f Political Science, 22, 233-249. Taylor, S. E., & Fiske, S. T. (1978). Salience, attention, and attribution: Top of the head phenome­ na. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 11, pp. 89-134). New York: Academic Press. Tetlock, P. E. (1983). Accountability and complexity of thought. Journal o f Personality and Social Psychology, 45, 74-83. Tetlock, P. E. (1986). A value pluralism model of ideological reasoning. Journal o f Personality and Social Psychology, 50, 819-827. Thaler, R. (1980). Toward a positive theory of consumer choice. Journal o f Economic Behavior and Organization, 1, 39-60. Thaler, R. (1987). The psychology and economics conference handbook. In M. W. Reder & R. Hogarth (Eds.), Rational choice: The contrast between economics and psychology (pp. 95-100). Chicago: University of Chicago Press. Tichenor, P. G., Donohue, T ., & Olien, C. (1970). Mass media flow and differential growth in knowledge. Public Opinion Quarterly, 34, 159-170.

3.

COGNITIVE PERSPECTIVE IN POLITICAL PSYCHOLOGY

187

Tichenor, P. G., Donohue, T., & Olien, C. (1980). Community conflict and the press. Beverly Hills, CA: Sage. Tobin, J. (1958). Liquidity preferences as behavior toward risk. Review o f Economic Studies, 25, 65-86. Tyler, T. R. (1980). Impact of directly and indirectly experienced events: The origins of crime-related judgments and behaviors. Journal o f Personality and Social Psychology, 39, 13-28. Weatherford, S. (1983). Economic voting and the symbolics politics argument: A reinterpretation and synthesis. American Political Science Review, 77, 158-174. Wilson, W. R. (1979). Feeling more than we know: Exposure effects without learning. Journal o f Personality and Social Psychology, 37, 811-821. Wyer, R. S., Jr. (1989). Social memory and social judgment. In P. Solomon, G. Goethals, C. Kel­ ley, & B. Stephens (Eds.), Memory: Interdisciplinary approaches (pp. 243-270). New York: Springer-Verlag. Wyer, R. S., Jr., Budesheim, T. L., Shavitt, S., Riggle, E., Melton, R. J., & Kuklinski, J. (1991). Image, issues, and ideology: The processing of information about political candidates. Journal o f Personality and Social Psychology, 61, 533-545. Wyer, R. S., Jr., & Gordon, S. E. (1982). The recall of information about persons and groups. Jour­ nal o f Experimental Social Psychology, 16, 128-164. Wyer, R. S., Jr., & Gordon, S. E. (1984). The cognitive representation of social information. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 2, pp. 73-150). Hills­ dale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., & Hartwick, J. (1980). The role of information retrieval and conditional inference processes in belief formation and change. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 13, pp. 241-284). New York: Academic Press. Wyer, R. S., Jr., & Martin, L. L. (1986). Person memory: The role of traits, group stereotypes, and specific behaviors in the cognitive representation of persons. Journal o f Personality and So­ cial Psychology, 50, 661-675. Wyer, R. S., Jr., & Ottati, V. (1992). Political information processing. In S. Iyengar & W. J. McGuire (Eds.), Explorations in political psychology (pp. 264-295). Durham NC: Duke University Press. Wyer, R. S., Jr., & Srull, T. K. (1981). Category accessibility: Some theoretical and empirical is­ sues concerning the processing of social stimulus information. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario Symposium (Vol. 1, pp. 161-198). Hills­ dale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., & Srull, T. K. (1986). Human cognition in its social context. Psychological Review, 93, 322-359. Wyer, R. S., Jr., & Srull, T. K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., Srull, T. K., Gordon, S. E., & Hartwick, J. (1982). The effects of taking perspec­ tive on the recall of prose material. Journal o f Personality and Social Psychology, 43, 674-688. Wyer, R. S., Jr., Strack, F., & Fuhrman, R. (1984). The acquisition o f information about persons: Effects o f task objectives and personal expectancies. Unpublished manuscript, University of Il­ linois, Champaign. Zajonc, R. B. (1984). On the primacy of affect. American Psychologist, 39, 117-123. Zaller, J., & Feldman, S. (1988, August). Answering questions vs. revealing preferences. Paper present­ ed at the fifth annual meeting of the Political Methodology Society, Los Angeles. Zucker, H. G. (1978). The variable nature of news media influence. In B. Rubin (Ed.), Communica­ tion yearbook (Vol. 2, pp. 225-245). New Brunswick, NJ: Transaction Books.

This page intentionally left blank

4

Some Cognitive Structures and Processes Relevant to Relationship Functioning

Margaret S. Clark Vicki S. Helgeson Kristin Mickelson Sherri P. Pataki Carnegie Mellon University

Contents Some Cognitive Structures Relevant to Relationship Functioning 190 Models of the Self and Others 190 Norms Governing the Giving and Receiving of Benefits 196 Satisfaction with, Commitment to, and Investment in Relationships 202 Organized Action Sequences 205 Prototypes of Emotions/Interpretations of Emotions 207 Some Cognitive Processes of Importance to Interpersonal Relationships 209 Social Comparison 209 Social Reflection and Further Evidence for Social Comparison 210 Processes Involved in Developing a Sense of Intimacy 214 Influences of One’s Own Emotional States on Perceptions of and Understanding of Others 216 Misattribution of Arousal 220 Attributional Processes 222 Relationship-Serving Biases 225 Summary and State of the Field 228 Acknowledgments 230 References 231

Psychological research in the area o f relationships has expanded rapidly since the 1970s. Most research can be considered cognitive in some sense, yet it is impossible to review it all. Thus, our strategy here is to review some important, but selected, topics. We have organized these topics into two broad groups. First, we talk about some cognitive concepts or structures that, relationship researchers have postulated, most people have stored in their minds and that have 189

190

CLARK E T A L .

important impacts on relationships. These include: (a) models of the self as a social object; (b) models of others as social objects; (c) norms governing the giving and receiving of benefits in relationships; (d) elaborate scripts regard­ ing how the behaviors of close others will interface with our own; and (e) con­ cepts such as trust, satisfaction, commitment, investment, and emotion in rela­ tionships. Second, we talk about some cognitive processes that appear to have important influences on the nature of relationships. These processes include the way affect influences people’s perceptions of and behavior toward others, attributional tend­ encies to evaluate close others’ behavior in positive or negative terms, processes leading people to feel that they have intimate relationships with others, social comparison, social reflection, and a variety of cognitive biases that appear help­ ful in maintaining relationships. Our goal is to convey a good sense of some of the exciting research on re­ lationships that has been conducted in recent years. Our division of topics into two broad groups—one on cognitive structures and one on cognitive processes— is really quite arbitrary. The structures discussed have considerable relevance to relationship processes. The processes discussed assume some important un­ derlying cognitive structures. Still, we utilize the distinction for heuristic pur­ poses. A final point prior to discussing this research is that, as we cover individual types of cognitive structures and individual cognitive processes relevant to rela­ tionships, the reader will surely note that one section does not flow or lead natur­ ally to the next. These discontinuities reflect the current state of research in the relationship field. A growing number of researchers in a wide array of laborato­ ries have been hard at work identifying cognitive structures and processes rele­ vant to relationships. However, they have been doing so largely independently of one another. Although they are generally well aware of others’ work, most researchers have not been building on each other’s work, nor trying to integrate the work that exists. Of course, integration is an important issue, and it is one to which we return at the end of the chapter.

SOME COGNITIVE STRUCTURES RELEVANT TO RELATIONSHIP FUNCTIONING Models of the Self and Others

Attachment Processes. As infants, people are born into a social world. They quickly form a conception of themselves as social objects. Are they persons toward whom others are attracted? Are they deserving of others’ care and attention? Are they socially valuable? They also quickly form a conception of what others in their social world are like. Are they to be trusted? Do they care about them?

4.

COGNITIVE STRUCTURES AND PROCESSES

191

Can they be counted on when they need them? Attachment theorists have sug­ gested that people develop general expectations about themselves and others based on early childhood experiences and that, once formalized, these beliefs are re­ sistant to change. They have argued that these expectations have important im­ plications for people’s close relationships throughout life. Until recently, the concept of attachment has been explored most extensively in relation to infants. Bowlby’s (1969/1982, 1973, 1980) pioneering research focused on infants’ emotional attachments to their primary caregiver, in most cases the mothers. He suggested that this relationship is unique as evidenced by a number of universal responses to caregivers among infants. First, infants and young children become distressed and tend to protest when separated from their primary caregivers, even when being well cared for by others. Second, as long as their caregivers are nearby, infants show an interest in exploring their environment and interacting with others. Third, when infants experience threat or discomfort, they cease exploration and seek out their caregivers for protection and comfort. From an evolutionary perspective, these responses have adaptive significance. They promote proximity to the caregiver, thereby ensuring the safety of the infant. The process of forming this unique relationship with the primary caregiver has come to be known as attachment, and the primary caregiver often is re­ ferred to as the attachment figure. According to Bowlby, through repeated in­ teractions with the attachment figure, children develop cognitive representa­ tions of themselves and their relationships to others, referred to by Bowlby (1973) as “internal working models.” These internal working models provide a basis for generalized expectations about the responsiveness and dependability of others. The original formulation of this theory has led to a vast amount of research with infants and young children. One critical focus of this research has been to examine individual differences in attachment experiences. Based on Bowlby’s work, Ainsworth, Blehar, Waters, and Wall (1978) developed a procedure for assessing infants’ attachment style by focusing on children’s reactions to separa­ tions and reunions with their primary caregivers. On the basis of this procedure, referred to as the Strange Situation, Ainsworth et al. delineated three patterns of attachment: secure, anxious/ambivalent, and avoidant. The secure pattern of attachment is believed to be the prototype for a wellfunctioning relationship. This is evidenced by children’s distress at separation from the attachment figure, comfort seeking upon reunion, and willingness to explore in the presence of the attachment figure. Primary caregivers of securely attached infants tend to be sensitive and responsive to their infants’ signals and needs (Ains­ worth et al., 1978). Anxious/ambivalent and avoidant patterns of attachment are believed to reflect less ideal forms of attachment and often are grouped together under the label “insecure.” Caregivers of children classified as anxious/ambiva­ lent respond inconsistently to their children’s needs. At times, the caregivers are

192

CLARK E T A L .

unavailable or unresponsive to the children. At other times, the caregivers are intrusive or overly demanding of the children’s attention. These infants tend to be preoccupied with their caregivers’ availability during the Strange Situation; they are distressed prior to separation, unable to be comforted upon reunion with the caregivers, and unwilling to explore in the caregivers’ presence. Caregivers of children classified as avoidant tend to avoid close bodily contact with their children and to reject their infants’ bids for proximity. These children exhibit lit­ tle overt distress during separation or reunion in the Strange Situation, and they tend to focus their attention on toys or other objects, rather than the caregivers. This reaction, often referred to as detachment, is believed to be an active, defen­ sive response to the mothers’ unavailability. Considerable thought and research have been put into understanding the processes and implications of infant-caregiver attachment for later development. Several empirical studies have indicated that individual differences in attachment classification during infancy predict various aspects of functioning during early childhood. For example, Waters, Wippman, and Sroufe (1979, Study 2) found that infants classified as secure at 15 months were more socially competent and ego-resilient at age 3.5 than those classified as insecure. Similarly, Erickson, Sroufe, and Egeland (1985) found that infants classified as insecure at 12 and 18 months were more likely than secure children to have behavior problems at ages 4.5-5. Secure children were rated higher than anxious/ambivalent children on measures of confidence, assertiveness, and social skills. Secure children also were rated as being less dependent on teachers and having better social skills than avoidant children. Teachers rated avoidant children as being more withdrawn and giving up more easily than securely attached children and more hostile than anxious/ambivalent children. However, there is some question, regarding the relative importance of early childhood interactions versus the importance of ongoing child care (Lamb, 1987; Ruble & Thompson, 1992). Lamb (1987), for instance, suggested the empirical evidence in support of the link between attachment during infancy and behavior during childhood is ambiguous. In particular, he noted that associations have been found only when the quality of child care was likely to have been stable. It may be that the quality of ongoing child care, rather than early parent-child interac­ tions, is responsible for the proposed link between early attachment and subse­ quent performance. Erickson et al. (1985) provided evidence that insecure children may, in fact, benefit from improvements in care and that secure children may suffer from a deterioration in care, but they maintained that early attachment ex­ periences should have a lasting impact beyond contemporaneous care. Specifi­ cally, they suggested insecure children should be particularly vulnerable to lapses in the quality of child care and secure children should rebound quickly following periods of impoverished care. Attachment theory also suggests that there should be some continuity between

4.

COGNITIVE STRUCTURES AND PROCESSES

193

infant-caregiver attachment and attachment experiences in adulthood. People with trusting and supportive relationships with attachment figures as children should, as adults, have models of relationships leading them to expect others to respond in a similar trustworthy manner. As a result, they should seek out others who meet their expectations. In contrast, those without dependable early relationships should lack such models, and instead have ones suggesting others may not be dependable. Anxious/ambivalent individuals may become excessively clinging and jealous in relationships, whereas avoidant individuals may avoid close rela­ tionships altogether. Such patterns of behavior may confirm and perpetuate be­ liefs about others. Implications o f Attachment for Adult Relationships. Hazan and Shaver (1987; Shaver & Hazan, 1992; Shaver, Hazan, & Bradshaw, 1988) paved the way for research on attachment among adults. They suggested conceptualizing adult roman­ tic love as an attachment process. In an initial study, they predicted adult roman­ tic relationships would evidence attachment-related behaviors that are similar to those observed in infant-caregiver relationships, such as proximity seeking and turning to the attachment figure for comfort. Second, they speculated that the patterns of attachment identified in infancy would be evident in adult relation­ ships. Third, they predicted that individual differences in adult attachment would be related to differences in childhood relationships with parents. They constructed a “love quiz” that asked readers of a newspaper questions about their most important love relationship. They included a three-item meas­ ure of adult attachment designed to reflect the three infant attachment styles delineated by Ainsworth et al. (1978). Hundreds of adults responded. Results indicated a distribution of attachment styles remarkably similar to that found in studies of infants and young children. Fifty-six percent of the respondents classi­ fied themselves as secure, 25% as avoidant, and 19% as anxious/ambivalent (com­ pared with approximately 62% of American infants classified as secure, 23% as avoidant, and 15% as anxious/ambivalent). Although all respondents reported some common, core features of romantic love, those who placed themselves into the three different attachment categories emphasized different things in connection with their most important relationships. Secure respondents described their most important relationships as happy, friendly, and trusting. Avoidant respondents described a fear of intimacy, emotional highs and lows, and jealousy. Anxious/ambivalent respondents described feelings of ob­ session, extreme sexual attraction, and jealousy. Similarly, respondents classifying themselves in the different ways also endorsed different beliefs about romantic relationships. For example, avoidant respondents were most likely to agree with statements such as “It is rare to find a person one can really fall in love with.” Finally, analyses of items designed to assess the subjects’ attachment history also indicated that subjects’ responses reflected their attachment classification in

194

CLARK ET AL.

meaningful ways. Secure subjects, in comparison with both avoidant and anx­ ious/ambivalent subjects, reported warmer relationships with their parents and warmer relationships between their parents. Insecure subjects recalled less favor­ able experiences. Avoidant subjects were most likely to describe their mothers as cold and rejecting, and anxious/ambivalent subjects were most likely to describe their fathers as unfair. The results of this survey provided preliminary support for the predictions based on attachment theory. Since this initial study, other researchers have reported similar findings based on different populations including subjects outside of the United States (Feeney & Noller, 1990; Mikulincer & Erev, 1991). Researchers have developed other measures of adult attachment, including the Adult Attach­ ment Interview, which consists of 18 open-ended questions (George, Kaplan, & Main, 1985). Among the many studies inspired by attachment theory, Collins and Read (1990, Study 3) focused on the importance of experiences with the opposite-sex parent during childhood on adult romantic relationships. They predicted a cor­ respondence between the style of attachment to one’s romantic partner and the caregiving styles of one’s parents, especially the opposite-sex parent, based on the idea that the opposite-sex parent provides a model for heterosexual relation­ ships. Again, this notion of continuity in relationships and expectations for rela­ tionships based on past experiences is based on the idea that people possess mental models of relationships, a central premise of attachment theory. In support of their prediction, Collins and Read found that only dating couples’ descriptions of their opposite-sex parents were related to the attachment style dimensions of their romantic partners. Specifically, women who perceived their fathers as hav­ ing been warm and responsive were more likely than women who saw their fathers as cold, distant, or inconsistent to be dating men who felt they could depend on others and who felt comfortable getting close. Men who described their mothers as cold or inconsistent were more likely to be dating women who were anxious, and men who described their mothers as cold and distant were more likely to be dating women who felt they could not depend on others. These results pro­ vide additional evidence that people seek out others who confirm their expecta­ tions about relationships. They also suggest that the opposite-sex parent may play a central role in shaping beliefs and expectations related to heterosexual rela­ tionships. Recently, Bartholomew (1990) proposed an expanded model of adult attach­ ment. It contains four attachment styles and assumes two underlying dimensions of adult attachment reflecting a person’s internal models of the self and others. Models of the self can be positive or negative (i.e., the self is worthy or unwor­ thy of love and attention). Views of others can be dichotomized similarly (i.e., others are trustworthy, caring, and available or rejecting, uncaring, and distant). The combination of these two dimensions forms the basis of four attachment categories (see Fig. 4.1). When a person has a model of the self as worthy, a

4.

COGNITIVE STRUCTURES AND PROCESSES

195

Models of the Self

Trustworthy Caring Available

Worthy

Unworthy

Secure

Preoccupied

Dismissing

Fearful

Model of Others Rejecting Uncaring Distant

FIG. 4.1. (1990).

An expanded model of attachment styles suggested by Bartholomew

model of the other as trustworthy produces a secure style; a model of the other as rejecting produces a dismissing style. When a person has a model of the self as unworthy, a model of the other as trustworthy produces a preoccupied style; and a model of the other as rejecting produces a fearful style. The secure and preoccupied styles are conceptually similar to Hazan and Shaver’s (1987) secure and anxious/ambivalent attachment styles. The fearful style is believed to be similar to the avoidant style described by Hazan and Shaver. However, the dismissing style reflects a distinct pattern of avoidant attachment. The dismissing person main­ tains a positive view of the self while holding a negative view of others. A dis­ missing person downplays the importance of close relationships and emphasizes the virtues of self-reliance and independence. By contrast, the fearful person main­ tains a negative view of both self and others. This person avoids close relation­ ships as protection from anticipated rejection. By distinguishing between these two avoidant styles, Bartholomew differentiates between those who avoid rela­ tionships due to lack of interest or motivation (dismissing) and those who avoid relationships due to fear of rejection (fearful). Bartholomew and Horowitz (1991) developed a four-item self-report meas­ ure, as well as an interview, to assess individuals’ attachment classification and have found that those who held a negative view of the self (the preoccupied and the fearful) reported high levels of insecurity, but only the fearful indicated less readiness to become intimate with and rely on other people. Those who held a negative view of others (the fearful and the dismissing) reported similar difficul­ ties in becoming close to and relying on others, but only the fearful had a low sense of self-worth. Undoubtedly, ongoing work in this area will continue to develop and refine attachment theory. In particular, good, but difficult, longitudinal work is still need­ ed to nail down the proposed long-term consequences of working models of the self and others formed in infancy and childhood. We also need more work on

196

CLARK E T A L .

the mechanisms by which these models may change over time. Recently, Collins and Read (in press) proposed that adult representations of attachment are best conceptualized as a hierarchical network of multiple, interconnected models. At the top of the hierarchy is an individual’s most general default model of relation­ ships that is based on an abstracted history of relationship experiences. Further down in the hierarchy are more specific models which apply to an increasingly narrow range of situations or specific relationships. Collins and Read suggest that this hierarchy develops over time and becomes increasingly complex as peo­ ple differentiate one relationship from another. Because all subsequent models are believed to be influenced by and to share common elements with models of earlier relationships, this cognitive structure accounts for the ongoing impact of early relationships, yet it also provides a basis for change as people modify their beliefs based on new experiences. The work reviewed here, as well as much other work that could not be reviewed, has argued that people do have cognitive models of themselves and others that have important implications for understanding relationships. Given the evidence that these models are formed early and may be relatively resistant to change, and given other, recent evidence that such models may be passed from generation to generation (Fonagy, Steele, & Steele, 1991) (because mothers’ cog­ nitive representations of childhood experiences with their parents predispose them to behave in particular ways toward their own children), it seems especially im­ portant to continue work in this area. Norms Governing the Giving and Receiving of Benefits

Another different type of cognitive structure that researchers have suggested is important to understanding relationship functioning is norms or rules governing the giving and receiving of benefits in relationships (e.g., Clark & Mills, 1979; Fiske, 1992; Lerner, Miller, & Holmes, 1976). Researchers studying such norms have assumed that people possess rules that guide behavior in social relationships and that may even account for how information about relationships is stored in memory (Fiske, Haslam, & Fiske, 1991). When the rules are followed, they facili­ tate smooth interaction between partners and promote and maintain positive feel­ ings between them. When they are violated, they disrupt behavior and cause negative feelings. Equity. A variety of such norms governing the giving and receiving of benefits have been proposed. For instance, some have suggested a rule of equity. Fol­ lowing an equity norm involves striving to keep the ratio of one’s inputs into a relationship relative to one’s outcomes equal to the ratio of the other’s inputs rela­ tive to his or her outcomes. It has been argued that this rule applies to all types of social relationships (Walster, Walster, & Berscheid, 1978). Much research has supported the application of this rule to relationships between strangers meeting

4.

COGNITIVE STRUCTURES AND PROCESSES

197

for the first time and to relationships in business settings (Walster et al., 1978). More recently, researchers reported some evidence consistent with a claim that it applies to more intimate relationships as well (Lloyd, Cate, & Henton, 1982; Sabetelli & Cecil-Pigo, 1985; Utne, Hatfield, Traupman, & Greenberger, 1984). For example, Utne et al. (1984) surveyed a large number of couples who were planning to marry. Males and females were interviewed separately. Perceived equity was assessed via two global self-report measures. Results showed that peo­ ple classified as following the equity norm scored slightly, but significantly, higher on measures of marital contentment and stability than did people who rated them­ selves as being under- or overbenefited. Sabatelli and Cecil-Pigo (1985) also obtained support for the use of an equity norm in intimate relationships. They found that reports of high overall equity in marriages were correlated with more favorable evaluations of outcomes de­ rived from marriages and with higher degrees of commitment to marriages. Desmarais and Lerner (1989) reported similar results among married couples and among undergraduates involved in dating relationships. Finally, Sprecher (1986) found that college students involved in intimate relationships who held global im­ pressions that their relationships were equitable experienced more positive and less negative affect in those relationships than did students who felt their rela­ tionships were inequitable. In a separate sample, she also found that perceptions of global equity were associated positively and significantly with reports of com­ mitment to the relationship (Sprecher, 1988). Even so, other evidence has suggested that an equity rule may not apply to close relationships such as friendships, romantic relationships, and family rela­ tionships. For instance, Lujansky and Mikula (1983) reported a study in which male students completed questionnaires about their girlfriends. Rather than meas­ uring equity globally, students rated the presence of 28 potential relationship in­ puts and outcomes, as well as the positivity and negativity of these inputs and outcomes. A measure of overall equity was calculated from these ratings. In this study, equity did not predict perceptions of the quality of the relationship, nor whether the relationship was intact 5 months later. Cate, Lloyd, and Henton (1985) observed that measures of global equity did not distinguish stable from unstable intimate relationships. Finally, some equity theorists have predicted that, “all things being equal,” relationships should become more equitable over time (Hatfield, Traupmann, Sprecher, Utne, & Hay, 1985; Hatfield, Utne, & Traupmann, 1979). Although some evidence for this has been reported (Schafer & Keith, 1981), in other studies such evidence nas not been obtained (Berg & McQuinn, 1986; Traup­ mann, Hatfield, & Sprecher, 1981). Indeed, at least one researcher actually ob­ served equity to decline over time. This occurred in a study of roommate relationships reported by Berg (1984). At this point, it seems safe to conclude that, although there is considerable and consistent evidence for equity applying to relationships between relative strangers, acquaintances, and business associ­ ates, the evidence for its applicability to intimate relationships is mixed.

198

CLARK E T A L .

Equality. Others have suggested that an equality rule is the rule that governs at least some intimate relationships. That is, each person should receive outcomes that are equal to the other person’s outcomes (Deutsch, 1975, 1985; Lemer, Miller, & Holmes, 1976). Interestingly, there has been evidence consistent with the use and value of this rule in intimate relationships as well. In one such study, Green­ berg (1983) had subjects read a story describing a lunch meeting between two people. One person ordered a large meal costing four times as much as the small snack ordered by the other. In one version, each person paid the exact amount of his or her meal. In a second version, they split the bill equally. In a third ver­ sion, no information was provided regarding how the bill would be divided. Sub­ jects judged diners who divided the bill equally as liking each other more, having a closer relationship, being better friends, and being more likely to see one another in the future, compared with diners who divided checks equitably or diners whose check division strategy was not described. Also, Austin (1980) predicted and found that when distributing rewards fol­ lowing a task, roommates (who presumably were fairly likely to be friends) tended to divide resources equally, regardless of whether they had contributed more or less to the task than had the other person. (In contrast, strangers tended to follow an equity, or what Austin called a merit rule, if they had contributed more than the other and an equality rule if they had contributed less than the other.) Needs-Based Rules and the Communal/Exchange Distinction. Still others have claimed that a needs-based rule governs the giving and receiving of benefits at least in intimate relationships. A needs-based rule dictates that people should benefit the other when that other has a need (Deutsch, 1975, 1985; Lamm & Schwinger, 1980, 1983; Mills & Clark, 1982). One group of researchers, Clark, Mills and their colleagues, have drawn a distinction between what they call communal and exchange relationships based on the norms governing the giving and receiving of benefits in each, and they have argued that these norms have wide-ranging effects on attraction and behavior in relationships (Clark & Mills, 1979; Mills & Clark, 1982). The communal norm is a cognitive structure presumably possessed by most members of society. It is frequently applied in relationships with friends, romantic partners, and family members. It dictates that people should be concerned with the other’s welfare and they should give benefits to that other when he or she has a need or simply to demonstrate a general concern for that other. A benefit given according to communal norm does not create a debt that the recipient must repay. Although some communal relationships are not mutual (e.g., a parent feels a strong responsibility for a very young child’s needs but does not expect the child to feel a similar responsibility for the parents’ needs), many are mutual. Under most circumstances, the communal norm dictates not only that one should respond to the other’s needs to the best of one’s abilities but also that the other should respond to one’s own needs to the best of his or her ability.

4.

COGNITIVE STRUCTURES AND PROCESSES

199

An exchange norm also presumably is possessed by most people in society. It is frequently applied in people’s relationships with strangers, acquaintances, and those with whom people do business. It dictates that benefits should be given in response to benefits received in the past or with the expectation of receiving comparable benefits in the future. Clark and her colleagues have accumulated a considerable amount of evidence supporting the idea that these norms exist in people’s minds and that they have an important impact on attraction and on behavior in relationships. For example, it has been found that when subjects are led to desire an exchange relationship with another, but not when they are led to desire a communal relationship, they like another who repays them for a benefit better than one who does not (Clark & Mills, 1979), they react positively to a request for repayment of a benefit received (Clark & Mills, 1979), and they feel exploited when another fails to repay them for a benefit given (Clark & Waddell, 1985). Moreover, members of exchange, but not members of communal relationships, strive to keep track of who has contributed what to a joint task for which there will be a reward (Clark, 1984). Offering more support for the fact that exchange norms do not apply to communal relationships, it has been found that when communal, but not exchange, relationships are desired, attraction drops when a repayment for a favor is received and when the other requests repayment for a benefit given (Clark & Mills, 1979), and that subjects will “bend over backwards” to avoid keeping track of who has contributed what to a joint task for which there will be a reward (Clark, 1984). Additional studies have shown these same sorts of differentiation in behaviors and reactions to others’ behaviors in naturally occurring communal and exchange relationships (Clark, 1984, Studies 2 and 3; Clark, Mills, & Corcoran, 1989). These findings support the existence of an exchange norm, which is applied selec­ tively to certain kinds of relationships. What about the existence of a communal norm? There is considerable evi­ dence for its existence as well. Studies have shown that people desiring or hav­ ing communal relationships pay more attention to the others’ needs (Clark et al., 1989; Clark, Mills & Powell, 1986) even when they can do nothing to meet those needs (Clark et al., 1986) than people desiring or having an exchange relation­ ship. They do this to a significantly greater extent than do those led to desire or naturally having an exchange relationship with the other. Moreover, people desiring a communal relative to those desiring an exchange relationship pay more attention to whether the other is putting effort into attending to their needs as well (Mehta, Clark, & Mills, 1992) and they help that other more (Clark, Ouellette, Powell, & Milberg, 1987). Finally, some recent research has shown that people desiring a communal relationship feel better after having helped the other than do people desiring an exchange relationship (Williamson & Clark, 1989, 1992). Thus, there is considerable evidence that people possess distinct, stored norms regarding the giving and receiving of benefits and that these norms have impor­ tant implications for their approaches toward relationships. However, these norms

200

CLARK ET AL.

or cognitive structures, are not quite as simple as they might seem at first. Clark and Mills pointed out that, although there is a qualitative distinction between com­ munal and exchange norms, there also is a quantitative dimension to communal norms (Mills & Clark, 1982; Mills, Clark, & Ford, 1992). That is, communal relationships vary in strength. People feel a stronger obligation to meet some peo­ ple’s needs (e.g., the needs of their child) than to meet other people’s needs (e.g., the needs of a friend). They will incur greater costs to meet the needs of those with whom they have stronger communal relationships. This idea implies the exist­ ence of still another cognitive structure influencing the nature of relationships. That is, people have an implicit hierarchy of communal relationships, with the needs of those with whom they have the strongest communal relationships taking precedence over the needs of those with whom they have weaker communal rela­ tionships. This hierarchy becomes an important determinant of behavior when making a decision regarding whose needs should be met, when to meet one per­ son’s needs would conflict with the ability to meet another person’s needs. For example, parents typically place their children’s needs above their friends’ needs. Thus, when faced simultaneously with meeting the need of a child (who needs a ride to soccer practice) and meeting the need of a friend (who needs a ride to the store), they are likely to meet the need of the child first. Attempting to Integrate a Confusing Literature. There is considerable sup­ port for the communal/exchange distinction. However, stepping back from this particular line of research, and considering the entire literature on norms governing the giving and receiving of benefits in relationships such as friendships, roman­ tic relationships, and family relationships, it may appear to the reader that the research on equity norms, equality norms, and communal norms is very mixed and that it is actually not very clear which norms apply to which relationships. Are the various programs of research on relationship norms that we described really as contradictory as they might seem at first? Clark and Chrisman (in press) suggest they are not. They take the position that it is a communal, or needs-based, rule that people believe is the ideal for friendships, romantic relationships, and family relation­ ships, and they argue that people do strive to live up to this rule. How do they account, then, for results that seemingly support use of equity or equality rules in such relationships? Considering evidence for equity first, they acknowledge it is possible that carefully following an equity rule over the entire course of a rela­ tionship could account for reports of global equity being associated with satisfac­ tion in such relationships. However, noting that members of existing friendships have been observed not to keep track of individual inputs into joint tasks (Clark, 1984; Clark et al., 1989), they express skepticism that this is what is really going on. Instead, they propose several other explanations. One possibility is that there are two, simultaneous consequences of following a communal rule of mutual responsibility for one another’s needs. One consequence is an increase in satis­

4.

COGNITIVE STRUCTURES AND PROCESSES

201

faction in the relationship. A second consequence is that usually (although not necessarily always) this results in more objectively equitable than inequitable rela­ tionships over the long run (assuming roughly equal needs of most partners). This could explain the association between people’s overall sense of equity and satis­ faction without assuming adherence to an equity rule on a day-to-day basis. An alternative possibility is that feeling satisfied (for whatever reason) leads one to an overall sense that things are fine with the relationship. This, in turn, leads to higher judgments of overall equity when subjects are asked. This, too, can explain an association between reports of global equity and satisfaction without assuming that people even think about global equity until explicitly asked. Still another explanation is that dissatisfaction (perhaps created by a partner’s lack of responsiveness to a person’s needs) leads people to calculate exactly who has con­ tributed what to the relationship and who gets what benefits from the relation­ ship. Such calculation may lead, in turn, to perceptions of global inequity (which may or may not be accurate). This process also could produce the observed posi­ tive correlations between perceptions of overall equity and relationships satisfac­ tion without assuming that an equity rule, rather than a needs-based rule, guides most people’s behavior in intimate relationships. In any case, the point is that the equity findings need not be considered at odds with the findings supporting the existence of a need based norm. Clark and Chrisman (in press) argue further that research suggesting that peo­ ple apply an equality norm to their intimate relationships need not be considered at odds with the idea that people hold and strive to adhere to communal and ex­ change norms. For instance, they suggest Greenberg’s (1983) finding that diners who split their bill evenly are more likely to be perceived as friends than are those who split their bill equitably, or those for whom no information is given, is not at odds with the existence of a communal norm applying to friendships. Given no information about differential needs of the diners, observers probably guessed needs to be equal; and when needs are equal, people following communal norms should split bills evenly. Similarly, they note that Austin’s (1980) finding of roommates being more likely than others to divide joint rewards equally is also compatible with communal norms if the observers who had no information about needs assumed equal needs. To test a needs-based rule versus an equality one within the paradigms used by Austin or Greenberg, a manipulation of needs should be included in the de­ sign of the studies. If, in a study such as Austin’s, equality is the preferred re­ ward allocation rule even when there is evidence of unequal needs, or if, in a study such as Greenberg’s, an equality rule leads to greater perceived friendship than use of a needs-based norm, then one might question whether a needs-based rule is the ideal most people hold for intimate relationships. Clearly, more empirical and theoretical work is needed to integrate the many findings that have emerged from the social psychological literature on norms governing the giving and receiving of benefits. However, one thing that is clear

202

CLARK ET AL.

is that people do possess such norms: They may have explicit awareness of these norms, or they may only have implicit awareness of them. In any case, these norms have important implications for their behavior in and satisfaction with their relationships. Satisfaction with, Commitment to, and Investment in Relationships

Other largely cognitive concepts which have recently received research attention include “satisfaction with,” “commitment to,” and “investment in” relationships. These concepts are central to an “investment” theory of relationships which Rusbult has developed in recent years (Rusbult, 1980a; 1980b; 1983; and see Rusbult & Buunk, in press, for a review of the theory and supporting work). Rusbult’s theory is based on Thibaut and Kelley’s (1959; Kelley & Thibaut, 1978) original interdependence theory but elaborates extensively on that model. Perhaps the most central cognitive construct of the theory is commitment. Ac­ cording to Rusbult commitment is a psychological state that represents one’s long­ term orientation toward a relationship. It includes feeling attached to a partner and feeling a desire to maintain a relationship, for better or worse. It is a subjec­ tive state that includes beliefs and emotional components and it influences a wide range of behavior in the relationship. What determines the level of commitment in relationships? Many factors. One is a person’s satisfaction with the relationship. Satisfaction is defined as the per­ son’s global affect regarding the relationship or attraction to the relationship. It concerns the degree to which the person believes his or her partner fulfills im­ portant needs. How is it determined? An individual assesses his or her outcomes from the relationship and evaluates them relative to his or her comparison level (Thibaut & Kelley, 1959), that comparison level being the level of outcomes the person feels he or she deserves. The comparison level is determined by such things as experiences in past relationships and observations of the relationships that similar others have. Importantly, Rusbult emphasizes, satisfaction is only one determinant of commitment level. Thus it is not necessarily as good a predictor of behaviors such as staying in or leaving relationships as many people intuitively think it ought to be. Commitment also is strongly influenced by people’s investments into relation­ ships. Investments refer to resources that have been committed to the relation­ ship. There are two general kinds of investments: intrinsic investments and extrinsic investments. Intrinsic investments are resources that are put directly into the relationship. They include such things as time, emotional effort, and self­ disclosures. Extrinsic investments occur when initially extraneous resources be­ come connected, inextricably, to the relationship. Extrinsic investments include such things as mutual friends, shared memories or material possessions, and ac­ tivities/persons/objects/events uniquely associated with the relationship. Once

4.

COGNITIVE STRUCTURES AND PROCESSES

203

invested, neither intrinsic nor extrinsic investments can be removed easily from the relationship. Just as greater satisfaction predicts greater commitment to rela­ tionships, so do greater investments. Still another important cognitive construct influencing commitment is CL-alt— the comparison level a person has for alternatives (including the option of remain­ ing alone). This concept was originally proposed by Thibaut and Kelley (1959) and refers to the lowest level of outcomes a person will accept given available alter­ native opportunities. CL-alt is the standard people use to determine whether to stay in a relationship or to leave it. The lower the CL-alt, the more likely a per­ son is to stay in a relationship. The theory suggests that the alternative options to the relationship will contribute importantly to commitment to the relationship, as do satisfaction with a relationship and investments into a relationship. Thinking about the various cognitive (and emotional) constructs which Rusbult discusses, one can make a number of predictions about relationships. For example, this model argues that individuals should be more satisfied with their relationships to the extent that those relationships provide high rewards and low costs and to the extent that the sum of rewards minus costs exceeds the individu­ al’s comparison level for the particular type of relationship—be it a romantic rela­ tionship, a friendship, or whatever. However, the model further predicts that measures of satisfaction alone often will not be sufficient for the development of commitment to relationships and thus will not necessarily predict staying in or leaving a relationship. One must also take into account a person’s alternative options and investment size. If, for instance, investment is high and good alter­ natives are not available, a person may feel committed to the relationship and stay in it regardless of extreme dissatisfaction with the relationship. The model also suggests that one may leave a relationship despite being satisfied. Why? Given low investments and good alternatives, one may feel little commitment and find it easy to move on to one of those available alternatives. Is there evidence to support Rusbult’s model and, consequently, the existence of the cognitive (and emotional) structures that she and others working in the interdependence tradition have postulated? The results from a long series of studies (e.g., Rusbult, 1980a; 1980b; 1983; Rusbult & Farrell, 1983; Rusbult, Johnson, & Morrow, 1986), Rusbult & Martz, 1992; Rusbult, Van Lange, & Verette, 1992 as reported by Rusbult & Buunk, in press; Rusbult, Verette, Whitney, Slovik & Lipkus, 1991) reveal that the answer is yes. For instance, in a longitudinal study of heterosexual dating relationships reported in 1983, Rusbult found that increases in rewards over time led to corresponding increases in reported satis­ faction. Further, commitment increased as a joint function of increases in satis­ faction, declines in the perceived quality of available alternatives, and increases in investment size. Greater rewards also promoted increases in commitment to maintain relationships. (The only parts of the model not receiving support were those relating costs to perceived satisfaction and commitment. Changes in costs generally did not affect commitment or satisfaction.)

204

CLARK ET AL.

Rusbult also found that the variables in her model helped to identify three types of people: stayers (people who remained in dating relationships over time), leavers (people who left their partners), and those who were abandoned (people whose partners left them). For stayers, rewards increased, costs rose slightly, satisfac­ tion grew, alternative quality declined, investment size increased, and commit­ ment grew—just as was expected from the model. The reverse trends were observed for leavers. The pattern of findings for those who were abandoned was perhaps most interesting. They showed few increases in satisfaction, but their alternatives declined in quality while they continued to invest heavily in their relationships—then their partners left. To give another example of research supporting the model, consider one recent study in which the model successfully predicted staying in abusive rela­ tionships versus leaving them. In this recent work, Rusbult and Martz (1992) examined the determinants of whether women who had been abused physical­ ly by their spouses or other partners and had sought safe haven at a women’s shelter returned to their partners soon after leaving the shelter or, rather, never returned or delayed returning for several months. Rusbult and Martz predicted that women’s satisfaction with their partners—in particular whether they re­ ported having any positive feelings left for them at all—would not be a very powerful predictor of returning. Why not? Because such feelings would be only one small determinant of feelings of commitment. Other things, such as investments into the relationship (e.g., having joint children, having more than one child) and the quality of alternatives, including the important option of staying alone (as indicated by such things as having money available, a job, transportation) also contribute to commitment (as measured by stated intents to return). Consequently, they should be important determinants of the likelihood of return. As predicted on the basis of the theory, but contrary to intuition, hav­ ing positive feelings (or a lack thereof) for the partners did not predict return. However, measures of commitment, investments, and the quality of alternatives did predict return—each in the manner suggested by the model. Finally, we note that Rusbult and her colleagues have reported evidence that high subjective commitment is associated with people believing that their relationship is superior to those of others (Rusbult et al., 1992 as reported by Rusbult & Buunk, in press). Commitment has also been shown to be related to people engaging in pro-relationship behaviors even when those behaviors may be costly or in opposition to direct self-interest. In particular, Johnson and Rusbult (1989) found that downplaying an alternative partner’s desirability is positively associated with high commitment to one’s relationships, and Rusbult et al. (1991) found that commitment plays a role in mediating persons’ willing­ ness to accommodate to their partner when that partner behaves poorly (i.e., be­ ing willing to inhibit impulses to react destructively and instead to react constructively).

4.

COGNITIVE STRUCTURES AND PROCESSES

205

Organized Action Sequences

Still another type of cognitive structure proposed by social psychologists that has important implications for relationships is an internal representation of organized action sequences (Berscheid, 1983; Kelley et al., 1983). This structure bears some similarity to Schank and Abelson’s (1977) earlier conception of a script. An or­ ganized action sequence is a sequence of behaviors that have been performed in succession on a repeated basis. These are well-practiced sequences that tend to “run off” automatically with little conscious thought. Berscheid (1983) suggested that much of the interaction taking place in rela­ tionships, particularly in intimate relationships of any duration, undoubtedly is made up of such organized action sequences. For example, a husband and wife who must get their children off to school each morning may learn a behavior sequence that is carried out each day and is executed with little variance or con­ scious thought. The husband may get up first and prepare the children’s breakfast while the wife gets showered and dressed. She then may wake the children and guide them downstairs to the already prepared breakfast. Her husband may hand her a cup of coffee, which she drinks as she prepares his lunch and he showers and changes. While he is doing this, she may be guiding the children in washing faces, brushing teeth, and getting dressed. When he finishes getting ready, he may collect the dressed children and drive them off to school. Due to the repetition and invariance of the sequence, over time it becomes represented internally and emitted as a single unit, rather than as a sequence of discrete acts. Once these sequences are learned, their execution requires very little, if any, conscious thought. With frequency of execution, the action system becomes an invariant and tightly organized cognitive structure, and once it be­ gins, it is very likely to follow through to completion. It is now difficult to modify. Organized action sequences also can occur over a considerable amount of time and can be organized into a hierarchical structure of plans and goals. The hus­ band and wife who diligently send their children off to school each morning also may be following a higher order plan to have well-educated children who will continue on to college someday. These higher order plans can vary in their de­ tail, level of organization, and stage of completion. Within a relationship, higher order plans may be based on shared goals to which each partner is committed. Over time, a couple’s organized action sequences can become tightly inter­ woven and highly interdependent. That is, there may be a large number of events in one partner’s behavior sequence that are connected causally to events in the other partner’s behavior sequence. As a result, completion of the action sequence is not possible without contributions from both people. In these cases, the par­ ticipant’s organized behavior sequences are said to be “meshed” or causally inter­ connected. Meshed sequences are similar to any other behavior sequence. With repetition and learning, they tend to show little variation, to follow through to completion, and to require little conscious thought.

206

CLARK ET AL.

These cognitive structures suggest a way of thinking about closeness in rela­ tionships. A close relationship can be characterized as one in which the causal interconnections between participants are frequent, strong, and diverse (Berscheid, Snyder, & Omoto, 1989; Kelley et al., 1983). As a relationship endures over time, the number of interconnections in these sequences may increase as a result of consistent behavior patterns. In fact, some relationships may consist almost entirely of such intertwined, organized sequences, involving both simple tasks (such as making dinner) and higher order plans (e.g., saving money to buy a house). It is important to note, however, that the length of a relationship does not neces­ sarily indicate that the participants’ behavior sequences will be highly intercon­ nected. Many long-term relationships consist of two people who can be described as having fairly separate individual sets of sequences, with few interconnections. They can be described as operating in parallel. In these relationships, the be­ havior and plans of one member have little impact on the behavior and plans of the other member. In such relationships, there is a relative lack of interdependence. Based on this conceptualization of interdependence, Berscheid et al. (1989) developed the Relationship Closeness Inventory (RCI) to assess the degree of in­ terdependence in a relationship. The RCI measures the frequency and diversity of activities engaged in by the two partners, as well as the extent to which each participant is influenced by the other person. Respondents are asked to indicate the total amount of time spent with their partners each day, the number of differ­ ent activities they have engaged in together over the past week, and the amount of influence their partners have on various aspects of their lives. Together, these three components are used to approximate the extent of interdependence in the relationship. Berscheid (1982, 1983) argued that the nature of these well-organized se­ quences, in particular the degree to which those of one person are meshed with those of the other, has important implications for relationships. She suggested that the amount of interconnectedness between two people’s chains is indicative of the degree of emotional investment in a relationship. When well-practiced be­ havior sequences run off smoothly, there may be little affect experienced by either partner. However, when well-practiced behavior sequences are interrupted, both members of the relationship may experience arousal, which is then likely to be labeled and experienced as emotion. Depending on the nature of the interruption, the experience of emotion can be either positive or negative. If the interruption interferes with completion of a behavior sequence, the emotional response should be negative. If the interrup­ tion facilitates the completion of a behavior sequence, the emotional response should be positive. For example, a wife who receives a phone call from her hus­ band informing her that he will be unable to stop at the grocery store after work may experience negative emotion, because this will interrupt the usual routine for making dinner. On the other hand, if she is told he will be unable to stop

4.

COGNITIVE STRUCTURES AND PROCESSES

207

at the store but has made reservations at a restaurant, she may experience posi­ tive affect. In this case, there still is an interruption but it facilitates the goal of getting dinner. The relationship between interdependence and the experience of emotion with­ in relationships has a number of interesting implications. Clearly, the level of interdependence in a relationship should predict distress experienced in the ab­ sence of one of the participants. That is, the greater the number of interconnect­ ed behavior sequences, the greater the amount of distress that should be experienced with the absence of the other person or with the dissolution of the relationship. Simpson (1987) provided empirical support for this prediction by measuring the degree of interdependence in premarital relationships using the RCI, satis­ faction with the relationship and later the amount of emotional distress experienced following a break-up. Interestingly, previous satisfaction with the partner was not related to distress following a breakup. Rather, individuals involved in rela­ tionships characterized by high interdependence experienced the greatest distress following dissolution of their relationship. Individuals who reported ease in find­ ing a suitable alternative, however, experienced less distress following the breakup. Perhaps the alternative partner alleviated the negative emotional response to the breakup by facilitating plans and routines that had been interrupted. Prototypes of Emotions/Interpretations of Emotions

Berscheid and her associates have not been the only ones to propose the exist­ ence of cognitive structures relevant to understanding emotion in relationships. Others in the relationship field have begun to explore how people interpret the emotion that partners bring to relationships. They have begun to ask what knowl­ edge people typically store about emotion and about its implications for their rela­ tionships. The work relevant to this has not all come from the same tradition. Prototypes o f Emotion. One relevant program of research has been conduct­ ed by Shaver, Schwartz and their colleagues (Schwartz & Shaver, 1987; Shaver & Schwartz, 1987; Shaver, Schwartz, Kirson, & O’Connor, 1987). According to their research, people have emotion prototypes stored in their memories. These prototypes include their implicit understanding of and beliefs about emotions. Prototypes were identified by cluster analyzing people’s descriptions of actual emotional experiences. Three components of the experiences of emotions such as anger, fear, sadness, joy, and love were identified: antecedents, responses, and self-control procedures. For example, people believe sadness begins with perceptions of loss, harm, or defeat, often in the realm of relationships. They also believe that sadness leads to responses such as crying, withdrawal, and reduced talking. Procedures are enacted to control the emotion in the case of sadness, including attempts to alter or eradicate existing circumstances through energetic

208

CLARK ET AL.

activity, denial, or optimism. The general point is that people seem to know a lot about typical behavior associated with emotions and, we assume, this has im­ portant implications for how they will react to people whom they perceive to be experiencing these emotions. For instance, because a sad person will be perceived as having experienced a loss of some sort, others should react to that person in predictable ways, by offering condolences or reassurances that the loss can be ameliorated. Given the perception that sad people withdraw, others also may choose to leave sad others alone for awhile—at least until the person seems ready to make some moves toward self-control. Still, the implications of the sorts of emotional scripts identified by Shaver and his colleagues for social relationships have yet to be explored thoroughi y .

Interpretations o f and Reactions to Others' Emotions. A few others also have asked how people react to information that another is feeling a particular way. For instance, Sommers (1984) asked respondents what emotions they and others experienced on typical days. Then she asked for observers’ reactions to the ex­ pression of these emotions. Positive emotions were judged to be more typical than negative emotions. Furthermore, people experiencing primarily positive af­ fect were seen as more sociable, conventional, popular, and likeable than others. Sommers also found that females who expressed negative affect were seen as more unsociable and unpopular than males who expressed the same affect, sug­ gesting, perhaps, that people have stored information in memory that females tend not to express as much negative affect as males, thereby leading to more extreme judgments of females who do express negative affect. However, when the negative affect is depression, there is some evidence that males are viewed more negatively than females (Hammen & Peters, 1977, 1978), perhaps because the emotion violates gender role expectations. Clark and Taraban (1991) found evidence that happiness increases people’s liking for others and that sadness causes others to be perceived as needy and de­ pendent. Their research also revealed that interpretations of expressed emotions not only depend on people’s concepts of what the emotion itself means but also on the type of relationship within which the expression of emotion takes place. In particular, Clark and Taraban found that expressions of happiness, sadness, and anger were all reacted to more positively in the context of communal rela­ tionships compared with exchange relationships. Indeed, although expressing anger significantly reduced liking in exchange relationships and sadness tended to do the same in exchange relationships, neither effect held in the context of a com­ munal relationship. Emotion may convey information about needs and express­ ing needs is viewed as more appropriate in the context of a communal than in the context of an exchange relationship. In any case, it seems clear that people have shared ideas about what particular emotions mean as well as ideas about when expressing those emotions is appropriate. These ideas undoubtedly have

4.

COGNITIVE STRUCTURES AND PROCESSES

2 09

important implications for how people behave in their relationships when faced with expressions of those emotions. Now that we have reviewed some cognitive structures relevant to understand­ ing relationships, we turn to a discussion of some cognitive processes relevant to relationship functioning. Again, we emphasize that our division of topics into these two sections is somewhat arbitrary. It already should be clear to the reader that some of the structures just discussed are closely related to cognitive process­ es. Similarly, the processes to be discussed next involve the assumption of some cognitive structures. Again, we emphasize that placing topics into these two categories was really a matter of judging whether particular research programs were more related to structure than to process or vice versa. SOME COGNITIVE PROCESSES OF IMPORTANCE TO INTERPERSONAL RELATIONSHIPS Two cognitive processes that have been shown by Tesser to be of considerable importance to understanding relationships are the processes of social comparison and social reflection. We start with social comparison—the process of compar­ ing oneself to others with respect to one’s abilities, opinions, and experiences. Social Comparison

Festinger (1954) originally proposed that people compare themselves with others to evaluate their opinions or abilities when an objective physical standard does not exist. In a series of manuscripts, Tesser addressed how this process relates to one’s self-esteem and to relationship functioning. (For a review of much of this work, see Tesser, 1988.) In particular, he noted that social comparison has the potential to influence one’s self-esteem. When one performs better than a similar other in a domain relevant to one’s identity, the comparison should result in a positive self-evaluation and self-esteem should rise. When one performs more poorly than a similar other, the comparison should make one evaluate oneself poorly and should decrease self-esteem. Tesser notes that the effects of social comparison processes on self-esteem should be greater the closer the relation­ ship or, in Tesser’s terms, the more a pair is seen as being in a unit relationship in a Heiderian sense. Thus, if earning a lot of money is important to my identity, and my sister who is similar in age and education earns more than I do, my self­ esteem should drop. It should drop more than it would have if my third cousin who is older and has more education earns more than I do. What are the implications of social comparison processes for relationships? We think there are many. The theory suggests some risk factors for distress in close relationships. That is, couples, friends, or siblings who are close and who are pursuing very similar goals should be at greater risk for distress than mem­ bers of relationships who are not pursuing similar goals. Moreover, the theory

210

CLARK E T A L .

suggests some strategies to prevent such distress. For instance, parents might be well advised to encourage their children to develop different interests, partic­ ularly if those children are close in age and of the same sex. Similarly, adults whose careers are important to them might be well advised to date other adults who are pursuing quite a different career or no career at all. Of course, other forces will work against such strategies. After all, often chil­ dren become interested in activities by virtue of seeing an older sibling partici­ pate in that activity. Also, because proximity is a powerful determinant of attraction (Festinger, 1951; Segal, 1974), it often will be the case that people pursuing a particular career path will become involved romantically with others pursuing the same career path or working in the same setting. Members of couples who are in the same career might be well advised to pursue different domains within that career or to draw finer distinctions between their work, or even to do all their work jointly so that social comparisons cannot be made. Is there evidence that social comparison processes influence relationships in the way Tesser predicts they will? The answer is yes. For example, the results of two studies reported by Tesser (1980) reveal that two people striving for good performance in the same domain produces more friction in close relationships. In particular, in one study of siblings, Tesser found that identification (i.e., per­ ceived similarity in interests, ways of thinking and feeling) decreased and fric­ tion (i.e., arguing) increased with increased closeness in age when the respondent reported that his or her sibling performed better on dimensions important to the respondents. In a second study, one based on autobiographies of eminent male scientists, evidence was found that the closeness of the relationship between fathers and their eminent scientist sons was a negative function of the similarity in their occupations. Social Reflection and Further Evidence for Social Comparison

A second part of Tesser’s self-evaluation maintenance theory has to do with “reflec­ tion processes.” Reflection refers to taking pride in another person’s positive characteristics. The reflection process raises self-esteem by allowing one mem­ ber of a relationship to bask in the other’s glory. As Tesser has noted in various articles, this process often involves claiming association with a well-known other. People frequently point out such things as that they live in the same neighbor­ hood as someone famous, have a cousin who qualified for the Olympic team, or are in the same academic department as someone who has won a Nobel prize. Of course, the reflection process does not require a person to explicitly point out an association with a positively valued other. Sometimes people benefit by reflection, simply because virtually everyone knows the association. For exam­ ple, Chelsea Clinton presumably does not need to point to her association with the president of the United States to derive the benefits of social reflection. The closer the relationship, the greater the effect of social reflection. For ex­ ample, the husband of a famous actress will experience a greater jump in self­

4.

COGNITIVE STRUCTURES AND PROCESSES

211

esteem than will the second cousin of the actress. On the other hand, the reflec­ tion process also can damage a person’s self-esteem if that person has the misfor­ tune of being associated with someone who has negative attributes or behavior. For instance, the relatives of a drug dealer whose activities have been outlined in detail on the front page of the newspaper may find that their self-esteem drops. This effect, too, would be exaggerated the closer the relationship. That is, the drug dealer’s parents will suffer more than the drug dealer’s nephew. What is the evidence that social reflection processes really operate? As Tesser has pointed out, some evidence for the operation of reflection processes predated his own work. Specifically, Cialdini and his colleagues reported some evidence for social reflection, or what they called basking in reflected glory (Cialdini et al., 1976; Cialdini & Richardson, 1980). In particular, Cialdini et al. (1976) con­ ducted a series of field experiments. In one study, they found that, on Monday morning, undergraduates at six universities were more likely to wear apparel that announced their university affiliation if the school football team had won a game the preceding Saturday. In another study, they found that students would use the pronoun we more frequently to describe the outcome of a football game if their team had won than if it had lost. Further, they found the tendency to emphasize a bond with successful others to be strongest after the students had failed a test publicly. This fact supports the assumption that drawing such associations bol­ sters students’ self-esteem. In a subsequent paper, Cialdini and Richardson (1980) reported two studies in which, after experiencing a visible personal failure, subjects enhanced the as­ serted quality of their home university, presumably to benefit from the reflected glory of the university. A second phenomenon was observed as well. That is, sub­ jects who had failed also devalued the asserted quality of a rival university, a phe­ nomenon the authors called “blasting.” Further support for both the reflection and blasting effects also has appeared in the literature (Cialdini & DeNicholas, 1989), and recently Hirt, Zillmann, Erickson, and Kennedy (1992) reported evidence for the consequences of social reflection on the “reflected upon” person’s predic­ tions of his or her own future performance. That is, they found that the fans of a university basketball team not only would predict that the team’s performance would be better in the future if the team won than if it lost, but also that their own performance at unrelated tasks would be better as well. Moreover, they found evidence that, although the team’s success influenced both mood and self-esteem among the fans, it was changes in the fans’ self-esteem, rather than in mood, that mediated fans’ estimates of both the team’s and their own future performance. Why should social reflection effectively boost self-esteem? Cialdini et al. (1976) suggested that Heider’s (1958) balance theory could provide the answers. Heider argued that people often strive to maintain cognitive harmony by thinking of things that are in a unit relationship with one another as also sharing an affective tone. This suggests that if people can associate with or cause themselves to be viewed as being in a unit relationship with a successful other or group, observers will come to view them more positively as well.

212

CLARK ET AL.

Of course, the work done by Cialdini et al. on reflection processes does not place it in the context of intimate or potentially intimate relationships, but some additional work by Tesser does. However, before describing this work, first note that the potential effects of reflection on relationships as described are diametri­ cally opposed to those of the comparison process as discussed earlier. That is, social comparison suggests that one may suffer as a result of another’s good per­ formance, and the closer the relationship the greater the suffering. In sharp con­ trast, social reflection suggests one may benefit as a result of another’s good performance, and the closer the relationship the greater the benefit. Thus, it is important to specify when social comparison and when social reflection will oc­ cur. According to Tesser, the key determinant of which process will apply in a given situation is the relevance of the performance domain to the person in ques­ tion. If the domain is not central to a person’s identity, the reflection process will dominate. If the domain is central to one’s self-definition, the comparison process will dominate. For example, if musical ability is absolutely irrelevant to a wom­ an, then, Tesser proposes, she can reap the benefits of social reflection without the pain of social comparison. That is, she may experience great pride and in­ creases in self-esteem when her sister consistently wins musical competitions and receives a large scholarship to a top music school. On the other hand, if the woman always has aspired to be a musician and has worked like crazy (but to not much avail) to get into a good music school, the sister’s good performance will cause her own self-esteem to drop and she is likely to feel miserable. Recall from the previous discussion that Tesser has asserted that both of these processes should have their most powerful effects the closer (in Heider’s sense) the rela­ tionship is. Certainly, friendships, romantic relationships, and family relation­ ships should be seen as being close or, in other words, as representing unit relationships. Therefore, Tesser and his colleagues assert that both social com­ parison and reflection should be most potent in these relationships. What is the evidence for this? One particular prediction that can be made about close relationships based on the theory (including both the ideas about reflection processes and social com­ parison processes) is that people will not always be more helpful to their friends than to strangers. Specifically, Tesser and Smith (1980) predicted that when task relevance is low, the reflection process should dominate and subjects should help the closer others. However, when task relevance is high, the comparison process should dominate and subjects should be reluctant to help closer others, because the comparison will make them look bad and lower their self-esteem. To test these ideas, Tesser and Smith had subjects perform a task high in relevance to them (it was said to relate to their verbal skills and leadership) or low in relevance to them (it was said to relate to very little about them). Subjects received feed­ back indicating they had performed poorly on the task and were then given a chance to give easy or difficult clues to either a friend or a stranger who had come to the study with them and who would be working on a similar task. As

4.

COGNITIVE STRUCTURES AND PROCESSES

213

predicted, when relevance was high, subjects gave more difficult clues to friends than to strangers. In contrast, when relevance was low, subjects gave more difficult clues to strangers than to friends. Another study supporting the model was reported by Tesser and Campbell (1982). As in the Tesser and Smith study, subjects participated in this study with a friend or a stranger. All subjects took a test of both social sensitivity and aes­ thetic judgment (and also rated the relevance of both skills to their own self­ definitions). Then, each subject received feedback on his or her performance. Next, they guessed how many right answers their friend and how many right answers a stranger had gotten on each test. As predicted by the theory, when the task was low in self-relevance, subjects were more charitable toward friends than strangers. That is, they were more likely to guess their friend had answered the question correctly. In contrast, when the task was relevant, just the reverse occurred—subjects were more charitable toward strangers than toward friends. In still another study, Pleban and Tesser (1981) showed that the theory could predict how much closeness a person will seek with another when the other out­ performs the person on a task. Male subjects first indicated how important vari­ ous domains of activity were to their self-definition. Then they and a stranger answered questions on a topic that was either high or low in relevance to the subject’s identity. The stranger did a little better, much better, a little worse, or much worse than the subject. Later, how close the subject chose to be to the stranger was measured. For high-relevance tasks, the greater the margin by which the other outperformed the subject, the farther away subjects sat from the stranger, the less desire subjects expressed to work with the stranger, and the less they said the stranger was “like” them. In contrast, if the domain was low in relevance, the greater the margin by which the other outperformed the subject, the closer subjects sat to the stranger, the more desire subjects expressed to work with the stranger, and the more subjects said the stranger was “like” them. Many other studies have supported Tesser’s self-esteem maintenance (SEM) model as well. For instance, it has been used successfully to predict what perfor­ mance dimensions people will consider relevant to them given the performance of others on those dimensions (Tesser & Campbell, 1980; Tesser & Paulhus, 1983). It also has proved useful in understanding some aspects of family dynam­ ics, such as siblings’ identification with one another and sibling rivalry (Tesser, 1980) and in understanding the occurrence of emotion in relationships (Tesser & Collins, 1988; Tesser, Millar, & Moore, 1988; Tesser, Pilkington, & McIn­ tosh, 1989). Interestingly, some recent work by Tesser and Cornell (1991) pro­ vides evidence that bolstering one’s self-esteem in ways completely separate from the SEM processes discussed here may “turn off’ the processes identified in the theory. That is, people whose self-esteem has been raised temporarily may be less subject to the effects of social comparison and reflection than are others. Before leaving the topic of social comparison, a few speculations about its importance to relationships that are not related to Tesser’s self-evaluation main­

214

CLARK ET AL.

tenance idea also might be made. Although Tesser has focused primarily on the self-evaluation function of social comparison, other researchers (not primarily in the relationships field) have discussed other functions of social comparison (Wood, 1989). For instance, some have discussed the possibility of social com­ parison serving self-improvement functions (Berger, 1977; Taylor & Lobel, 1989). A person may compare skills to a superior other with the goal of obtaining im­ portant information about what can be done to improve those skills (Berger, 1977). Others have pointed out that people may compare to enhance self-esteem (Crocker & Schwartz, 1985; Crocker, Thompson, McGraw, & Ingerman, 1987; Wills, 1981). That is, if a person fails a task, is diagnosed with a serious illness, or encounters some other misfortune, making downward comparisons to people who are even less fortunate may reduce distress and/or enhance self-esteem. We speculate that these functions of social comparison, like the self-evaluation function, also may prove to have important implications for the formation and functioning of a relationship. Ambitious, goal-oriented people may seek relation­ ships with accomplished people in their field to model themselves after those others. A misfortune may cause people to seek out relationships with other peo­ ple who have fallen victim to the same misfortune (perhaps in support groups, perhaps elsewhere) to be able to engage in downward comparisons and, as a result, to feel better. To date, the implications of the self-improvement and self­ enhancement functions of social comparison for understanding ongoing or poten­ tially intimate relationships has not been explored, but we believe that doing so holds great promise. Processes Involved in Developing a Sense of Intimacy

A process presumably important to overall relationship development is that un­ derlying the development of a sense of intimacy within that relationship. Although many researchers have assessed the intimacy of relationships, past research sel­ dom has discussed the process through which intimacy may develop. In a recent paper, Reis and Shaver (1988) attempted both to synthesize a diverse literature on intimacy and to propose a set of processes through which it may develop. Their model is interpersonal and transactional in nature. They propose that in­ timacy can be understood best as a process, rather than as a static state. The process begins with one person, an actor, disclosing or expressing some aspect of him or herself to the other person. This is the stage that has been the focus of most prior research. Reis and Shaver argue that this disclosure or expression may be verbal or nonverbal in nature. They also suggest that expressions of feelings, rather than factual information about the self, are more likely to foster intimacy. Following a disclosure or expression by one person, the partner then responds to that disclosure or expression, and both people take note of this response. Un­ like prior research, Reis and Shaver contend that the partner’s response to the actor is just as important in facilitating a sense of intimacy as is the actor’s initial

4.

COGNITIVE STRUCTURES AND PROCESSES

215

disclosure or expression, thus emphasizing the interpersonal nature of the process of achieving intimacy. According to the Reis/Shaver model, both partners approach the interac­ tion with their own needs, goals, desires, and fears regarding intimacy. These motives not only affect what one person discloses or expresses to the other, but also how the other responds to this information. In addition, the partner interprets the actor’s behavior and the actor interprets the partner’s response through an “interpretative filter.” These filters are composed of schemas and expectations based on past experiences. For example, a person with a history of nonresponsive partners may interpret the present partner’s initial silence as rejection, rather than as an attempt to listen and understand. The final deter­ minant of whether an interaction leads to a sense of intimacy lies in the actor’s reaction to his or her partner’s response. The actor must feel understood, vali­ dated, and cared for. Feeling understood stems from the actor’s perception that the partner accurately understands the disclosure or expression. Validation stems from the actor’s perception that the partner values and appreciates the actor. Feeling cared for stems from the partner’s expression of affection and/or emotion. A feedback loop is included in the model, such that the actor’s reaction to the partner’s response will affect the actor’s subsequent motives and fears, as well as the actor’s future disclosures and expressions. Thus, actors who feel under­ stood, validated, and cared for will be less fearful of self-disclosure and more likely to self-disclose to partners in the future. This model of intimacy processes is useful for describing intimate interactions or the development of intimacy within a relationship, but is not sufficient for describing intimate relationships. An intimate relationship involves more than just an accumulation of intimate interactions. Reis and Shaver (1988) suggest that intimate relationships develop their own emerging properties apart from those stemming from intimate interactions. These properties include: (a) a sense of “weness,” (b) a temporal perspective (past and future), (c) feelings of commitment that may override periodic negative feelings, (d) mutual responsivity, and (e) the development of trust based on a shared history of positive experiences. Nonethe­ less, the intimacy process seems to be central to close relationships and it may be that intimate interactions contribute to the emerging properties. When college students were asked about the importance of a number of activities in friendship, four of the five most highly rated friendship goals involved the components described in Reis and Shaver’s model of intimacy: (a) having someone with whom to share personal thoughts and feelings, (b) having a few special friends who care about you, (c) having someone who understands you, and (d) having someone to whom you can tell your deepest secret (Reis, 1990). These stated goals are similar to the self-disclosure, feeling cared for, and feeling understood aspects of Reis and Shaver’s model. This model of intimacy processes is more recent than other models of rela­

216

CLARK E T A L .

tionship processes discussed in this section. Thus, it remains relatively untested. However, it does hold considerable promise for fruitfully guiding future research. Influences of One's Own Emotional States on Perceptions of and Understanding of Others

Impact o f Emotions on Judgments o f Others. The general area of research on the role of emotion in social relationships has received considerable attention lately. We already have touched on this area twice: once when we discussed how the structure of interdependence in relationships as it is represented in the par­ ticipants’ minds can influence the emotions experienced in relationships, and a second time when we discussed how emotion prototypes and perceptions of other’s emotions might influence relationships. Another topic that falls under the category of emotion and relationships is the question of whether the emotions that people bring with them to relationships (or that arise during the course of interactions) influence those relationships, and, if so, through what processes. We begin with the question of how the emotional state that a single individual brings to a relationship influences that relationship. In general, research in this area has indicated that people in good moods tend to view the world and other people in a more positive manner and to behave toward others in a friendly or positive manner. These perceptions and behaviors may then serve to maintain their pleasant mood. In contrast, people in bad moods have been shown to have a more pessimistic view. That is, they have been shown to see the negative side of events and people, although some have noted the research findings on the ef­ fects of negative mood have been generally less consistent than the findings on the impact of positive moods. To give some concrete examples, Bower (1981) has found that subjects hyp­ notized to feel angry were more likely than subjects made to feel happy to inter­ pret pictures of other people in a negative manner. Similarly, Forgas, Bower, and Krantz (1984) have found that subjects hypnotized to feel good perceived that both they and a partner behaved more positively than did subjects hypno­ tized to feel bad. Further, Forgas and Bower (1987) found that when presented with both positive and negative information about another person, happy sub­ jects formed more favorable impressions and made more positive judgments than did sad subjects. Subjects also paid more attention to mood consistent informa­ tion (i.e., positive details when in a good mood and negative details when in a bad mood) and were more likely to later recall mood-consistent information. Additional studies have attempted to assess the separate effects of positive and negative mood on person perception by including neutral mood control groups. For instance, in a recent study, Erber (1991) induced positive, neutral, or nega­ tive moods by exposing subjects to brief stories with happy, neutral, or sad con­ tent. Then, in a supposedly separate study, subjects read about some people who were each described by a positive and a negative trait category (e.g., moody and

4.

COGNITIVE STRUCTURES AND PROCESSES

217

warm). Next, subjects rated the likelihood that those people would behave in ways applicable to the positive trait and the negative trait. Relative to what occurred in the neutral mood (control) group, positive moods increased the perceived likeli­ hood of the people behaving in ways consistent with the positive trait; negative moods increased the perceived likelihood of the people behaving in ways consis­ tent with the negative trait. (Interestingly, in this study, estimates of positive and negative behaviors unrelated to those traits were not influenced.) Erber replicat­ ed the effects for negative mood in a separate study in which mood was manipu­ lated using a facial feedback technique. These studies, taken together with quite a few others, provide evidence for positive moods increasing the positivity of judgments about others (Clark & Wad­ dell, 1983; Fiedler, Pampe, & Scherf, 1986; Forgas & Moylan, 1987) and some evidence for negative moods decreasing the positivity of judgments about others (Griffitt, 1970), social interest in others (Crandall, 1978), and the perceptions of the amount of social support available from others (Procidano & Heller, 1983). Still other studies suggest that people’s general level of arousal influences how they perceive others. For example, when presented with ambiguous facial ex­ pressions, people tend to attribute not only their own emotional state, but also their level of arousal, to others (Schiffenbauer, 1974). Also, individuals in a high state of arousal following exercise are more likely than individuals in a low state of arousal to interpret a positive facial expression as indicating joy, a high arousal positive emotion, than serenity, a low arousal positive emotion (Clark, Milberg, & Erber, 1984). A few boundary conditions on the general finding that positive moods bias people’s perceptions of themselves and others in a positive direction, whereas negative moods bias such perceptions in a negative direction have been observed. For instance, positive moods have not been found to enhance impressions of others about whom there is simply little positive information altogether, such as judg­ ments regarding drunk drivers and heroin traffickers (Forgas & Moylan, 1987). Also, the arousal that accompanies strong positive emotions sometimes may be misattributed to feelings of dislike for unattractive partners. That is, positive moods, if accompanied by noticeable arousal and if occurring close in time to exposure to a very unattractive partner, may be misattributed to feelings of dis­ gust. Similarly, the arousal that accompanies strong negative emotions sometimes may be misattributed to feelings of liking for attractive partners. That is, nega­ tive moods, if accompanied by noticeable arousal and if occurring close in time to exposure to a very attractive target, may be misattributed to feelings of attrac­ tion (White, Fishbein, & Rutstein, 1981). (This is discussed in more detail in the next section on misattribution of arousal.) In other words, the nature of the target person appears to have important implications for the effects that moods have on judgments of that person. It may be that transient mood states associated with relatively low levels of arousal are most likely to affect perceptions of others toward whom people have quite mixed feelings (Clark & Isen, 1982).

218

CLARK ET AL.

Possible Mechanisms. What mechanisms underlie such effects? Several have been proposed but a full discussion of them is beyond the scope of the present chapter. However, we will simply note some explanations and refer the reader to the original reports for fuller discussions of these issues. One explanation is that moods may increase the accessibility of mood-congruent thoughts stored in memory about the person to be judged (Bower, 1981; Clark & Isen, 1982; Isen, 1975; Isen, Shalker, Clark, & Karp, 1978). A second possibility is that mood cues broad, general styles of behaving. For example, positive moods might make people generally more sociable, agreeable, and cheerful. Negative moods might make people generally less sociable, more disagreeable, and gloomy (Cunningham, 1988b). Such general response styles could result in overall response tendencies to evaluate people and objects in a positive or negative manner without necessarily assuming moods cue individual pieces of target related, mood-congruent material from memory. A third explanation is that moods can serve as pieces of information that may be taken into account when making judgments (Schwarz & Clore, 1983, 1987). The very fact that one is in a particular mood might be attributed, correct­ ly or incorrectly, to something about the people present in one’s environment at the time the mood is experienced. Such attributions, in turn, might influence judgments about those people. Still anther possibility for why positive (but not for why negative) moods influence judgments is simply that people may wish to maintain their positive moods. Thus, they may try to perceive the best in others, to make positive interpretations of others’ behavior, and so forth in an effort to maintain or not to spoil their positive moods. There is some evidence consistent with all these views (see Clark & Williamson, 1989, for a review) but, as already mentioned, discussing them in detail goes beyond the scope of this chapter. What implications will mood-biased perceptions of other people have on satis­ faction with and behavior in relationships? This question has been less well researched than has the question of whether there are biases in the first place. However, we do suspect important implications exist. For instance, it seems straightforward to guess that people who see their partners in more or less posi­ tive ways also may react more positively or negatively toward those partners, and may interpret those partners’ actions more positively or negatively. Their emotionally colored view of the other also may influence the nature of interac­ tions or situations they might be willing to enter into with their partner (Kelley, 1984). Of course, the other may react more positively or negatively in return. In other words, we suspect the positive or negative moods that people bring with them to their relationships, especially if experienced chronically, may lead to selffulfilling prophecies, such that the initially mood-biased negative or positive im­ pressions of others may lead people to behave toward others in such a way as to confirm the initial impressions (cf. Snyder, Tanke, & Berscheid, 1977) even

4.

COGNITIVE STRUCTURES AND PROCESSES

2 19

though the initial positive or negative impression may be caused by factors com­ pletely separate from the relationship even from the perceiver’s point of view. It is interesting to note, in light of these speculations, that moods indeed have been shown not only to influence perceptions of other people but behavior rele­ vant to social interaction as well. For example, those in a negative mood have been shown to be motivated to withdraw from social contact, which inadvertent­ ly may perpetuate their negative mood (Cunningham, 1988a). Other, consistent research has indicated that people who are in a good mood are more interested in social activities (and in activity in general) than are people in a depressed mood. People who are in a depressed mood are more interested in sitting and thinking, being alone, and taking a nap (Cunningham, 1988b). In addition, people in nega­ tive moods have been shown to be more likely to deplete resources from a pool of goods in a manner that ultimately will hurt them (Knapp & Clark, 1991)—a sort of behavior that may be detrimental to their ability to cooperate successfully with others. It also has been demonstrated amply that those in positive moods are more likely to help others (Isen, 1970; Isen, Clark, & Schwartz, 1976) and more likely to self-disclose to a partner (Cunningham, 1988a) than are those not in a good mood. Perhaps these behavioral effects are due to mood biasing perceptions of others, and perhaps not. The answer to that awaits further research. In any case, be­ cause self-disclosure predicts satisfaction in dating relationships (Berg & McQuinn, 1986) and friendship formation (Cohen, Sherrod, & Clark, 1986), and because helping is likely to do the same, this work provides further support for the idea already expressed that moods brought to a relationship from the outside or aris­ ing within the relationship itself may set off interactive chains of events —often consistent in tone—that will have important effects on whether relationships are initiated, maintained, or deteriorate and perhaps even dissolve. Having said all this, one final point to make is that the gloomy portrait of the consequences of people bringing negative moods to their relationships just point­ ed out is a bit overstated. It is overdone because we also know that people some­ times fight the naturally negative biases that seem to arise from negative moods, perhaps in an effort to alleviate those negative moods—a strategic process described by Clark and Isen (1982). For instance, we know that people in negative moods have been shown in some studies to be more likely than those not in negative moods to behave in prosocial ways (Cialdini, Darby, & Vincent, 1973). Cialdini and his colleagues have suggested that the motivation for prosocial behavior (among adults who have been socialized to believe helping is good) often may be the alleviation of negative moods (Cialdini et al., 1973; Cialdini & Kenrick, 1976). Such mood-linked strategies may be conscious or they may be so well practiced and automatic that they are not conscious. In either case, they provide an optimistic note regarding the potential relationship relevant effects of the nega­ tive moods people bring to their relationships.

220

CLARK ET AL.

Misattribution of Arousal

Still another cognitive process with important implications for relationships, es­ pecially for establishing feelings of romantic loye, involves the misattribution of arousal to feelings of love. In the early 1970s, Berscheid and Walster (1974; Walster & Berscheid, 1971) proposed a theory of passionate love. At the time, a reinforcement paradigm was the dominant framework within which to view attraction. However, Walster and Berscheid found this paradigm inadequate for purposes of explaining certain aspects of passionate love. For instance, it seemed unable to explain such things as the excitement of new romances dying down over time, people still feeling in love with someone who has just dumped them, or the intensified feelings of romantic love between members of a couple when parents object to the relationship (Driscoll, Davis, & Lipitz, 1972). They specu­ lated that a two-component theory of love based on Schachter’s (1964) theory of emotion would be more useful for understanding these feelings of passion. Schachter and his colleagues had argued that the experience of emotion re­ quires that a person be both physiologically aroused and able to label or interpret that arousal in emotional terms (Schachter, 1964; Schachter & Singer, 1962). Building on those ideas, Berscheid and Walster (1974) proposed that feelings of passionate love arise from being physiologically aroused and from simultaneous­ ly having an appropriate target available to which the label of love can be attached. The arousal involved might come from a number of sources. It might be a direct result of something having to do with the other person. A person might feel nervous about whether an attractive, opposite sex other will be cordial if approached or will reject any advances. The presence of another person may elicit feelings of sexual arousal. Alternatively (or in addition), a person may be aroused in another’s presence for reasons having little to do with the other person. For instance, a person may have just exercised, ridden a roller coaster, or had an argument with another person. As long as the arousal is labeled (or mislabeled as the case may be) as romantic love, it will be experienced as such. The circumstances under which people will label arousal as romantic love presumably are influenced by culture. American culture teaches people it is ap­ propriate to be romantically attracted to single people about their own age who are physically attractive. This encourages people to experience feelings of romantic love in the presence of such others. The culture also teaches people that it is in­ appropriate to feel certain emotions in certain circumstances, something that may encourage mislabeling of arousal as romantic love on occasion. For instance, the culture may teach boys (but not girls) that they should not feel fear. Consequent­ ly, a male and a female may both experience arousal on a roller coaster ride. However, as a result of cultural norms, the male may be prone to mislabel those feelings as romantic love toward the female next to him. The female, feeling it is more permissible to feel fear, may be less prone to such misattribution of arousal, and thus less prone to experience romantic feelings in this particular situation.

4.

COGNITIVE STRUCTURES AND PROCESSES

221

A number of studies have supported these ideas about passionate love. In one, Dutton and Aron (1974) studied young adult males who were crossing one of two bridges in a park in Canada. One bridge, chosen to be the experimental, arousal producing, bridge, was a long, narrow suspension bridge made of wooden boards attached to cables that ran from one side of a canyon to the other. It was selected for its qualities, including “a tendency to tilt, sway and wobble,” very low handrails, and a 230-foot drop to rocks and shallow rapids below. The other, control, bridge was a solid wood bridge only 10 feet above a small, shallow rivulet. After each male crossed one of these bridges, a male or female experimenter stopped him and said he or she was conducting a study on the effects of scenic attractions on creative expression. The subject was then asked to write a brief story, was given the experimenter’s phone number, and was told that he could call the experimenter later if he wanted more details about the study. The content of the stories was scored later for sexual imagery, and the number of actual phone calls made to the experimenter was recorded. It was assumed that each of these things indicated attraction toward the experimenter. The results supported the theory. Only when there was reason for subjects to be aroused (i.e., they were on the suspension bridge) and an appropriate target (label) for romantic love was available (i.e., the experimenter was a female) did a substantial proportion of subjects call for more information (50% called). Moreover, only in those same conditions did subjects write stories containing considerable sexual imagery. In the other conditions (experimental bridge/male experimenter and control bridge/male or female experimenter), the percentages of subjects making phone calls and the sexual imagery scores were much lower. Two additional experimental studies supported the same ideas (Dutton & Aron, 1974). Considerably later, White et al. (1981) conducted a series of studies both to confirm the Berscheid/Walster theory of passionate love and to extend their ideas to understanding feelings of repulsion. White and his colleagues argued that arousal not only would be attributed to feelings of love if an appropriate, appealing tar­ get was available, but also would be attributed to repulsion if an appropriate, unappealing target was available. In one of their studies, male subjects volun­ teered for a study supposedly on the relationship between attitude similarity and initial attraction. Subjects were told to give their reactions to a variety of stimuli and that later they would interact with another, opposite sex, participant on a brief “date.” Subjects then were exposed to one of three audiotapes. Two of the tapes were designed to elicit arousal, one through exposure to negative stimuli (a grisly description of a killing and mutilation that the victim’s family watched) and one through exposure to positive stimuli (a comedy tape). The remaining tape was selected to be neutral both in affective tone and a tendency to elicit arousal. Pretests indicated that the two arousal tapes did heighten arousal and the control tape did not. Next, subjects saw a videotape of their partner/future date. Half of the sub­

222

CLARK ET AL.

jects in each arousal condition (and in the control condition) saw a quite attrac­ tive female, and half saw a quite unattractive female. Finally, they rated their attraction toward that other in a variety of ways, including rating how attractive her traits were, how attractive she was in general, and how attractive she was romantically. All the measures supported the hypotheses. When the potential fe­ male partner was attractive, arousal, produced either by the positive tape or by the negative tape, enhanced attraction relative to reported attraction in the control condition. In contrast, if the potential female partner was unattractive, arousal again produced either by the positive tape or by the negative tape resulted in lower attraction relative to the control condition. Apparently, Berscheid and Walster’s (1974) theory of passionate love can be extended to be a theory of repulsion as well. Attributional Processes

Most of the early research on attribution theory focused on attributions made for the behavior of strangers. More recently, researchers have begun to examine at­ tributional processes in the context of relationships (Baucom, 1987). Specifical­ ly, researchers have studied the attributions spouses make (i.e., the explanations spouses give) for events that occur in their relationships. Attributions typically are studied by presenting spouses with real or hypothetical marital events and asking them to rate these events on a variety of causal dimensions. Research results have led to the conclusion that spouses who are dissatisfied with their marital relationships make attributions that minimize the impact of their partner’s positive behavior and augment the impact of their partner’s negative be­ havior (Bradbury & Fincham, 1990a; Fincham & Bradbury, 1991). Distressed spouses make defensive attributions (i.e., they place more blame on their partner and less blame on themselves; Baucom, 1987). Causal Attributions. In studying the attributions spouses make for marital events, research has drawn on the three dimensions identified by Weiner et al. (1972): locus (internal, external), stability (stable, unstable), and globality (global vs. specific). Bradbury and Fincham (1990b) hypothesized that distressed spouses would attribute their partner’s negative behavior to internal, stable, and global causes. Research supports this prediction and concluded that, among the three attribution dimensions, attributions to global causes most reliably characterize distressed spouses (Baucom, Sayers, & Duhe, 1985; Fincham, 1985; Fincham & O’Leary, 1983). In general, distressed spouses appear to make more attributions than non­ distressed spouses (Bradbury & Fincham, 1990a), which may be explained by a greater frequency of negative behavior in distressed than in nondistressed rela­ tionships. Negative behavior promotes the attributional search because: (a) it is unexpected and (b) it is salient and disruptive (Baucom, 1987). Thus, to the ex­ tent that negative behavior is more common among distressed than nondistressed

4.

COGNITIVE STRUCTURES AND PROCESSES

223

couples, one would predict a greater attributional search among the former than the latter. Attributions for positive marital events are related less reliably to marital dis­ tress than are attributions for negative marital events (Baucom, Sayers, & Duhe, 1989; Fincham, Beach, & Nelson, 1987). Bradbury and Fincham (1990a) hypothe­ sized that distressed spouses would make attributions for their partner’s positive behavior to external, unstable, and specific causes. Marital dissatisfaction has been associated with the attribution of positive behavior to unstable causes and to specific causes, but not necessarily to external causes (see Bradbury & Fin­ cham, 1990a; for a review; Fincham & O’Leary, 1983). External attributions for marital events can be ambiguous; external may mean the spouse, the marital rela­ tionship, or causes outside the marital relationship. Internal attributions for posi­ tive relationship events (i.e., giving credit to one’s partner) are related more clearly to relationship satisfaction (Jacobson, McDonald, Follette, & Berley, 1985; Kelley, 1979; Thompson & Kelley, 1981). Relative Attributions. Attributions that spouses make for their partners’ be­ havior must be considered in relation to attributions spouses make for their own behavior (Bradbury & Fincham, 1990a; Fincham & Bradbury, 1991). By evalu­ ating relative attributions, previously inconsistent results, especially with regard to the locus dimension, often are clarified. For example, Camper, Jacobson, Holtzworth-Munroe, and Schmaling (1988) found that distressed spouses saw the relationship as a greater cause of negative behavior than the self, whereas non­ distressed spouses saw the self as a greater cause of negative behavior than the relationship. Similarly, Fincham (1985) found that distressed spouses perceived the cause of marital problems to be located more in the partner than in the self, and nondistressed spouses perceived the cause of marital problems to be located more in the self than the partner. Thus, it appears that distressed spouses make self-enhancing attributions, whereas nondistressed spouses make partner-enhancing attributions (Bradbury & Fincham, 1990a; Fincham & Beach, 1988; Fincham, Beach, & Baucom, 1987). It is interesting that self-enhancing attributions typi­ cally are related to better psychological adjustment for the individual, but appear to reflect worse adjustment in the context of a relationship. Responsibility Attributions. More recently, researchers have distinguished responsibility attributions from causal attributions, concluding that responsibility attributions may be the more important of the two in distinguishing distressed from nondistressed couples (Fincham & Bradbury, 1991). There is less doubt about the cause of a behavior in a marriage than about whether a person who behaved in a cer­ tain way intended for the behavior to have its effect. A behavior intended to have the effect that it had is a behavior for which one must be held responsible (Shaver, 1985). Evidence has been provided that spouses distinguish between causal and respon­ sibility attributions. Fincham and Bradbury (1992) found that responsibility

224

CLARK ET AL.

attributions, but not causal attributions, were associated with spouses’ expres­ sions of anger and whining when discussing marital problems. Responsibility at­ tributions include judgments of intentionality, motivation, and behavior justification. For negative behavior, distressed spouses perceive greater negative intentions, perceive more selfish motivations, and assign greater blame compared with nondistressed spouses, whereas for positive behavior, nondistressed spouses perceive more positive intention, deem behavior as more unselfishly motivated, and give more praise compared with distressed spouses (Fincham, Beach, & Nel­ son, 1987). Explanations for Attributions. There are two primary explanations for the differences between the attributions distressed and nondistressed spouses make for their partners’ behavior. First, the disparity in reports may reflect accurate appraisals of different behaviors. It is difficult to determine the accuracy of spouses’ attributions. Observers’ ratings cannot be used to evaluate accuracy, because ob­ servers are unfamiliar with the history of the relationship. Thus, the question as to whether distressed and/or nondistressed spouses are accurate in explaining their partner’s behavior remains an unresolved issue (Bradbury & Fincham, 1990a; Fincham & Bradbury, 1991). Second, the different explanations that distressed and nondistressed spouses give for marital events may be due to motivational biases, in particular to differen­ tial needs to protect self-esteem. In marriage, spouses not only are concerned with the image of themselves, but also with the image of their partners and their relationships (Baucom, 1987). The need for self-enhancement may not be as strong among nondistressed as among distressed couples, because spouses in nondis­ tressed couples gain a sense of self-esteem and security from the partner-enhancing attributions that occur (Baucom, 1987). Causality. The question arises as to whether attributions for marital events lead to distress or whether distress determines the attributions spouses make for marital events. The correlational nature of much of the research just discussed leaves the issue of causality unresolved. However, Bradbury and Fincham (1990a) cite three sources of evidence for the hypothesis that attributions cause marital satisfaction. First, in one study, results indicated that manipulating explanations for being in a relationship affected perceptions of love (Seligman, Fazio, & Zanna, 1980). Second, clinical outcome studies have shown that changes in attribu­ tions lead to changes in marital satisfaction. Third, longitudinal studies that assess attributions prior to relationship satisfaction show that attributions lead to changes in relationship satisfaction, but changes in relationship satisfaction do not affect later attributions. If attributions exert a causal impact on marital satisfaction, what is the process by which this effect occurs? Bradbury and Fincham (1990a) suggest that behavior mediates the relation of attributions to marital satisfaction. Spouses’ attributions

4.

COGNITIVE STRUCTURES AND PROCESSES

225

for marital events affect their behavior. Four studies have supported this idea. In one, responsibility attributions (but not causal attributions) were related to be­ havioral intentions (Fincham, Beach, & Nelson, 1987), such that when a part­ ner’s behavior was perceived as negatively intended, selfishly motivated, and blameworthy, the spouse intended to emit a punishing response. In a second study, spouses who perceived their partner’s negative behavior as intentional subsequently exhibited negative behavior. In turn, partners who perceived the spouse’s nega­ tive behavior as intentional reciprocated in kind (Bradbury, Fincham, & Beach, 1989). In a third study, spouses’ causal and responsibility attributions for marital difficulties were related to the occurrence of negative behavior when discussing marital difficulties (Bradbury & Fincham, 1990b). Finally, in a fourth study, at­ tributions for negative marital events were manipulated prior to discussion of such events and found to affect the behavior of distressed, but not of nondistressed, spouses (Fincham & Bradbury, 1988). Relationship-Serving Biases

In the domain of relationships, several biases have been detected that may have implications for the maintenance of those relationships. These biases include: (a) the previously described biased attributional accounts of relationship experiences, (b) selective memory for relationship-relevant events, and (c) biased perception of a partner’s attributes and abilities. We touch briefly on the second but devote most of this section to discussion of the third area of research. Selective Memory. Beach and Tesser (1988) suggest that commitment to a relationship increases the amount of time spent thinking about one’s partner. Think­ ing of a person about whom one’s thoughts are initially positive will lead to in­ creased positive thoughts; thinking about a person about whom one’s thoughts are initially negative will lead to increased negative thoughts. This process oc­ curs because of the addition of beliefs that are consistent with one’s present thoughts and the reinterpretation of previous beliefs to be consistent with one’s present thoughts (Tesser & Cowan, 1975). Biased Perceptions o f Partner. Under conditions of social interdependence, Rosen, Fiske, and Rosenthal (1991) argue that the ability to form an accurate impression of one’s partner may be impaired. For example, a supervisor’s ac­ curate evaluation of an employee’s job performance may be compromised and biased in a positive direction if that supervisor is involved with the employee romantically. To test this notion, these researchers examined the judgments that college students made of a confederate who gave either a competent or an in­ competent role-playing performance. The expectation of social interaction with the confederate was manipulated. The competence of the confederate’s task per­ formance was perceived as intended by subjects who did not expect to interact

226

CLARK ET AL.

with the confederate (i.e., the competent performance was evaluated more favor­ ably than the incompetent performance). In contrast, subjects who expected to interact with the target rated an incompetent task performance as positively as a competent task performance. Thus, accuracy in judging task competence was affected by social interdependence. These results were replicated in a second ex­ periment, in which a stronger manipulation of social interdependence was em­ ployed (half of the subjects were led to expect a date with the confederate). The authors concluded that social interdependence increases one’s motivation to see a potential partner in a positive light. There are other conditions in which negative behavior is cast in a positive light in the context of relationships. As discussed in the previous section on attributional biases in relationships, one way that an individual can maintain a good rela­ tionship when confronted with a partner’s negative behavior is to attribute that behavior to the situation (e.g., it was not his fault that he was late, his boss al­ ways gives him last minute assignments). But Murray and Holmes (1992) re­ cently pointed out that an accumulation of negative behaviors (e.g., he is late again and again) makes it difficult to avoid a dispositional attribution, and they suggested a different way partners can maintain a good impression of one another in this circumstance. Specifically, one way a person can remain committed to a relationship when faced with negative attributes in a romantic partner is to cre­ ate virtues where none seemed to exist before. To test this idea, Murray and Holmes (1992) conducted two laboratory ex­ periments with dating couples. In the first experiment, subjects were asked to indicate the extent to which their partners initiate conflict. Half of these subjects were then told that engaging in conflict promotes intimacy (experimental group), and half were not (control group). These researchers found that later the ex­ perimental subjects who initially perceived their partners as seldom initiating con­ flict changed their perceptions of their partners as initiating more conflict. In the second experiment, subjects were asked to describe similarities and differences between themselves and their partners. Half of these subjects were then told that the awareness of differences promotes intimacy (experimental group) and half were not (control group). Results indicated that experimental subjects recast similarities as differences (e.g., we come from similar backgrounds, but within those backgrounds there are important differences). When asked to write narra­ tives describing the development of intimacy in their relationships, experimental subjects in Study 1 were more likely to emphasize conflict and how conflict benefit­ ed the relationship, and experimental subjects in Study 2 were more likely to em­ phasize the importance of differences and how differences fostered intimacy (each as compared with control subjects). These studies suggest that when confronted with negative information about romantic partners, one way to maintain a positive view of the relationship is sim­ ply to recast the partner’s behavior to see what one wishes to see. Just as individuals seem capable of deriving the positive out of a stressful life event (e.g., cancer;

4.

COGNITIVE STRUCTURES AND PROCESSES

227

Taylor, 1983), relationship partners seem able to see positive attributes in their partner when they wish to see them. In the context of relationships, engaging in this process has been suggested by Brickman (1987) to do more than maintain the relationship—to actually increase feelings of romantic love. One remaining question is: What happens if or when positive reinterpretations of a partner’s be­ havior conflicts with reality? Another open question has to do with whether such reconceptualizations of partners’ behaviors actually might be harmful in some cases. If, when people perceive negative traits in their partners, they quickly dis­ tort what they “know” to see the partner in a more positive light, might that not tend to keep them in negative or abusive relationships that are destined to end at some point anyway? Biased Perceptions o f Non-Partners. A third relationship-serving bias centers on the evaluation of others, rather than on the evaluation of one’s partner. Ac^ cording to Johnson and Rusbult (1989), an attractive alternative is a major threat to relationship stability. They proposed that one mechanism by which people main­ tain their relationships is derogation of attractive alternatives. To test this, they conducted a study in which undergraduates were shown advertisements, some of which depicted members of the opposite-sex. Results indicated that subjects in more committed relationships perceived the opposite-sex people as less physi­ cally and sexually attractive than subjects in less committed relationships. Rus­ bult (1983) also reported another study supporting the idea that members of relationships will derogate alternatives. Results from this longitudinal study showed that increased involvement in a relationship was associated with describing poten­ tial alternatives in more negative terms (Rusbult, 1983). Results from a second longitudinal study showed that subjects viewed potential alternatives more nega­ tively over a 7-month period, but that this perception interacted with relationship status. Subjects who remained together at the end of 7 months devaluated alter­ natives, but subjects who had broken up enhanced alternatives (Johnson & Rus­ bult, 1989, Study 1). These results raise the question of whether a devaluation of alternatives is occurring among the more committed or an enhancement of alternatives is occurring among the less committed. If derogation of attractive alternatives is a relationship-maintenance mechanism, such derogations only should occur for the more realistic threats (Johnson & Rus­ bult, 1989; Simpson, Gangestad, & Lerma, 1990). To test this hypothesis, Simpson et al. (1990) showed college students slides depicting either older opposite-sex, young (similar in age to the subjects) same-sex, or young opposite-sex persons with the expectation that subjects who were in more committed relationships would derogate only the more realistic threats—the young opposite-sex persons (Simp­ son et al., 1990, Study 2). Results indicated that, as expected, exclusive daters evaluated the young opposite-sex others less favorably than the nonexclusive daters, but no differences occurred between the two groups’ evaluations of the old opposite-sex or young same-sex others. In a second study, subjects who

228

CLARK ET AL.

participated in a computer dating service were presented with low, moderate, and highly attractive targets, and the potential to date one of the targets was manipulated (Johnson & Rusbult, 1989, Study 2). Results indicated that subjects in more committed relationships rated the highly attractive target less favorably than did subjects in less committed relationships. In contrast, relationship com­ mitment did not affect the evaluation of low and moderately attractive targets. In addition, the devaluation of the attractive target was more likely to occur un­ der conditions in which the subject perceived a potential to date the target. The authors conclude the impact of commitment is greater under conditions of high threat to the relationship (i.e., the potential to date a highly attractive partner). This is consistent with other research on cognitive distortions that has suggested such distortions may increase under conditions of more severe threat (Helgeson, 1992; Taylor et al., 1992). Why might biased perceptions of one’s partner or of others be adaptive in the context of relationships? First, positively distorted views of a potential partner provide positive expectations that can lead to self-fulfilling prophecies and be­ havioral confirmation (Snyder et al., 1977). For example, if one expects one’s partner to be friendly, one may behave in a more pleasant manner, leading one’s partner to actually be more friendly. One’s expectation not only has been con­ firmed, but this expectation has influenced the partner’s behavior, which could have long-lasting implications. Second, derogating attractive alternatives may make the task of relationship maintenance easier. One is more likely to remain in a relationship to the extent to which one’s comparison level for alternatives falls short of one’s present relationship (Thibaut & Kelley, 1959).

SUMMARY AND STATE OF THE FIELD What Has Been Included in This Review and What Has Not? What we have done in this chapter is to provide brief overviews of various researchers’ pro­ grams of work on some cognitive structures and some cognitive processes rele­ vant to understanding what are known commonly as close or intimate relationships. All the work reviewed could be (and has been elsewhere) discussed in far more depth. We urge readers interested in particular topics to pursue them beyond the bounds of this chapter. There is also much work on cognitive structures and mechanisms of great relevance to relationships that we have not discussed at all due to space constraints. For example, one important cognitive construct for relationships that we have not discussed is trust. Holmes and his colleagues recently formulated a very in­ teresting model of this construct (Holmes, 1991; Holmes & Rempel, 1989). To give another example, an important cognitive process for relationships that has not been discussed here is emotional communication. This topic recently has received attention by quite a number of different relationship researchers (Buck,

4.

COGNITIVE STRUCTURES AND PROCESSES

22 9

1989; Gaelick, Bodenhausen, & Wyer, 1985; Gottman & Levenson, 1988; Noller, 1980; Noller & Ruzzene, 1991). Again, we urge the interested reader to go be­ yond this chapter to read about other cognitive structures and processes relevant to understanding relationships. Finally, we emphasize that other general approaches to describing work on cognition in close relationships exist. For instance, a recent book edited by Fletcher and Fincham (1991) entitled Cognition in Close Relationships includes chapters on cognition in close relationships written by a number of authors more closely associated with the social cognition perspective in social psychology than those represented in this review. In addition, Berscheid (in press), in the most recent Annual Review chapter covering interpersonal relationships, describes and dis­ cusses a great deal of research on the role of cognition in relationships that has not been included here. Reading Fletcher and Fincham’s (1991) book and Berscheid’s (in press) review in conjunction with the present chapter will provide the reader with a very comprehensive view of cognition in relationships. Do the Various Programs o f Research Reviewed Here Relate to One Another in Any Way? We conclude with a few comments about how the research pro­ grams reviewed in this chapter might relate to each other. Clearly, collectively, social and clinical psychologists have made considerable progress in coming to understand some cognitive structures and processes important to understanding close relationships. However, the research described in any one of our various subsections also has tended to come out of one or two laboratories. The researchers whose work has been described are generally well aware of one another’s work. Even so, the programs of research have not been interrelated with each other. Does this mean these programs of research do not relate in any way to each other? No. It is our strong belief that these programs of work interrelate in many ways. What are some likely connections? We will mention just a few. One thing we suspect is that the sorts of early models of relationships dis­ cussed by attachment theorists are closely related to many of the other constructs and processes we have discussed. For example, feeling securely attached— including having positive models of the other as someone who meets one’s needs and of the self as someone worthy of having one’s own needs met—is probably closely related to both the development and the application of communal norms to relationships. After all, learning that one’s needs are met consistently by a caregiver should form the very basis of a child’s understanding of the communal norm. Moreover, although most people probably have the concept of a communal relationship, securely attached people may be far more likely than others to be­ lieve that they and others can and will live up to communal norms. As a result, they, more than others, may readily apply communal norms to new relationships and end up forming more communal relationships than others. Similarly, it seems self-evident that early models of the self and of the other emerging from attachment processes should relate closely to concepts discussed

230

CLARK ET AL.

by Kelley, Thibaut, and Rusbult in connection with their interdependence the­ ory. For instance, those early models should strongly influence comparison lev­ els. People with more positive models of others and the self should set higher comparison levels, should be less satisfied in relationships that do not meet high standards, and should be less likely to stay in poor relationships or invest much in them. Finally, still in connection with attachment styles, consider the process­ es of social comparison and reflection in relationships discussed by Tesser. Who is going to be most subject to the effects of social comparison and of social reflec­ tion on self-esteem? As mentioned earlier, Tesser and Cornell (1991) noted that the self-esteem maintenance processes they have discussed may be turned off when people’s self-esteem is bolstered in other ways. Who is likely to be high in self­ esteem? It seems likely that those who are most securely attached—that is, those with positive models of themselves—will have higher self-esteem than others. Thus, one might predict that securely attached persons will be less subject than others to the sorts of cognitive processes discussed by Tesser and his colleagues. (On the other hand, it may be that people who are anxious-avoidant are especial­ ly likely to feel arousal in their relationships and, consequently, through the sorts of misattribution processes discussed by Berscheid and Walster, they may be more likely than others to experience frequent and intense feelings of romantic love.) Of course, it is not just the work on attachment styles that can be tied to other work. Many other connections between programs of research can be drawn. For example, when are the sorts of intimacy processes described by Reis and Shaver most likely to occur? These processes involve self-disclosure (which often in­ volves revealing one’s needs) and positive responsiveness to such disclosures. When will people care about each other’s needs and engage in intimacy process­ es? The communal/exchange work suggests it is in communal relationships that people feel a special responsibility for the other’s needs, and thus intimacy process­ es should occur far more often in the context of communal than exchange rela­ tionships. Further intimacy processes and the sorts of interdependence that Berscheid describes as indicating emotional investment in relationships both re­ late closely to Rusbult’s more general concept of investment in relationships. The greater the intimacy achieved and the greater the emotional investment, the greater should be a general sense of investment and the greater should be the commit­ ment to the relationship. We easily could go on, but we hope that the general point has been made ade­ quately. That is, although the programs of work described in this chapter on cog­ nitive structures and cognitive processes relevant to relationships may seem largely independent of one another, they need not be viewed that way. The entire field of relationship work is wide open to efforts at integration. We suspect that ef­ forts in this direction, as well as efforts aimed at investigating additional cogni­ tive structures and processes relevant to relationships, will prove quite fruitful and interesting.

4.

COGNITIVE STRUCTURES AND PROCESSES

231

ACKNOWLEDGMENTS Preparation of this chapter was supported in part by National Science Founda­ tion Grant BNS-9021603 and by a James McKeen Cattell sabbatical award to Margaret Clark.

REFERENCES Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns o f attachment: As­ sessed in the strange situation and at home. Hillsdale, NJ: Lawrence Erlbaum Associates. Austin, W. (1980). Friendship and fairness: Effects of type of relationship and task performance on choice of distribution rules. P ersonality and Social Psychology Bulletin, 6, 402-408. Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal o f Social and Personal Relationships, 7, 147-178. Bartholomew, K., & Horowitz, L. (1991). Attachment styles among young adults: A test of a fourcategory model. Journal o f Personality and Social Psychology, 61, 226-244. Baucom, D. H. (1987). Attributions in distressed relations: How can we explain them? In S. Duck & D. Perlman (Eds.), Intimate relationships: Development, dynamics, and deterioration (pp. 177-206). London: Sage. Baucom, D. H., Sayers, S. L., & Duhe, A. (1985). Assessing couples’attributions fo r marital events. Unpublished manuscript, University of North Carolina, Chapel Hill. Baucom, D. H., Sayers, S. L., & Duhe, A. (1989). Attributional style and attributional patterns among married couples. Journal o f Personality and Social Psychology, 56, 596-607. Beach, S., & Tesser, A. (1988). Love in marriage: A cognitive account. In R. J. Sternberg & M. L. Barnes (Eds.), The psychology o f love (pp. 330-355). New Haven, CT: Yale University Press. Berg, J. H. (1984). The development of friendship between roommates. Journal o f Personality and Social Psychology, 46, 346-356. Berg, J. H ., & McQuinn, R. D. (1986). Attraction and exchange in continuing and noncontinuing dating relationships. Journal o f Personality and Social Psychology, 50, 942-952. Berger, S. M. (1977). Social comparison, modeling, and perseverance. In J. M. Suls & R. L. Miller (Eds.), Social comparisons processes: Theoretical and empirical perspectives (pp. 209-234). Washington, DC: Hemisphere. Berscheid, E. (1982). Attraction and emotion in interpersonal relations. In M. S. Clark & S. T. Fiske (Eds.), Affect and cognition: The seventeenth annual Carnegie Symposium on Cognition. Hills­ dale, NJ: Lawrence Erlbaum Associates. Berscheid, E. (in press). Interpersonal relationships. In M. R. Rosenzweig, & L. W. Porter (Eds.), Annual review o f psychology. Palo Alto, CA: Annual Reviews, Inc. Berscheid, E. (1983). Emotion. In H. H. Kelley, E. Berscheid, A. Christensen, J. H. Harvey, T. L. Huston, G. Levinger, E. McClintock, L. A. Peplau, & D. R. Peterson (Eds.), Close rela­ tionships (pp. 110-168). New York: Freeman. Berscheid, E., Snyder, M ., & Omoto, A. M. (1989). The relationship closeness inventory: Assess­ ing the closeness of interpersonal relationships. Journal o f Personality and Social Psychology, 57, 792-807. Berscheid, E., & Walster, E. (1974). A little bit about love: Foundations o f interpersonal attraction. New York: Academic Press. Bower, G. (1981). Mood and memory. American Psychologist, 36, 129-148. Bowlby, J. (1969/1982). Attachment and loss: Vol. 1. Attachment. New York: Basic Books. Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New York: Basic Books.

232

CLARK ET AL.

Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression. New York: Basic Books. Bradbury, T. N., & Fincham, F. D. (1989). Behavior and satisfaction in marriage: Prospective medi­ ating processes. In C. Hendrick (Ed.), Review o f personality and social psychology (pp. 119-143). Newbury Park, CA: Sage. Bradbury, T. N., & Fincham, F. D. (1990a). Attributions in marriage: Review and critique. Psycho­ logical Bulletin, 107 , 3-33. Bradbury, T. N., & Fincham, F. D. (1990b). The impact o f attributions in marriage: A behavioral analysis. Unpublished manuscript. Bradbury, T. N., Fincham, F. D ., & Beach, S. R. (1989). The im pact o f attributions in marriage: Attributions and behavior exchange in marital satisfaction. Unpublished manuscript. Brickman, P. (1987). Commitment, conflict and caring. Englewood Cliffs, NJ: Princeton-Hall. Buck, R. (1989). Emotional communication in personal relationships: A developmental-interactionist view. In C. Hendrick (Ed.), Close relationships (pp. 144-163). Newbury Park, CA: Sage. Camper, P. M ., Jacobson, N. S., Holtzworth-Munroe, A., & Schmaling, K. B. (1988). Casual attri­ butions for interactional behaviors in married couples. Cognitive Therapy and Research, 12, 195-209. Cate, R. M ., Lloyd, S. A., & Henton, J. M. (1985). The effect of equity, equality and reward level on the stability of students’ premarital relationships. Journal o f Social Psychology, 125, 715-725. Cialdini, R. B., Borden, R. J., Thome, A., Walker, M. R., Freeman, S., & Sloan, L. R. (1976). Basking in reflected glory: Three (football) field studies. Journal o f Personality and Social P sy­ chology, 34, 366-375. Cialdini, R. B., Darby, B. L ., & Vincent, J. E. (1973). Transgression and altruism: A case for hedonism. Journal o f Personality and Social Psychology, 34, 907-914. Cialdini, R. B., & DeNicholas, M. E. (1989). Self-presentation by association. Journal o f Personal­ ity and Social Psychology, 57, 626-631. Cialdini, R. B., & Kenrick, D. T. (1976). Altruism as hedonism: A social development perspective on the relationship of negative mood state and helping. Journal o f Personality and Social Psychol­ ogy, 34, 907-914. Cialdini, R. B., & Richardson, K. D. (1980). Two indirect tactics of image management: Basking and blasting. Journal o f Personality and Social Psychology, 39, 406-415. Clark, M .S . (1984). Record keeping in two types of relationships. Journal o f Personality and Social Psychology, 47, 549-557. Clark, M. S., & Chrisman, K. (in press). Resource allocation in intimate relationships: Trying to make sense of a confusing literature. In M. Lem er & G. Mikula (Eds.), Entitlement and the Affectional bond. New York: Plenum. Clark, M. S., & Isen, A. M. (1982). Toward understanding the relationship between feeling states and social behavior. In A. H. Hastorf & A. M. Isen (Eds.), Cognitive social behavior (pp. 73-108). New York: Elsevier/North-Holland. Clark, M. S., Milberg, S., & Erber, R. (1984). Effects of arousal on interpretations of others’ emo­ tions. Journal o f Personality and Social Psychology, 46, 551-560. Clark, M. S., & Mills, J. (1979). Interpersonal attraction in exchange and communal relationships. Journal o f Personality and Social Psychology, 37, 12-24. Clark, M. S., Mills, J., & Corcoran, D. (1989). Keeping track of needs and inputs of friends and strangers. Personality and Social Psychology Bulletin, 15, 533-542. Clark, M. S., Mills, J., & Powell, M. C. (1986). Keeping track of needs in communal and exchange relationships. Journal o f Personality and Social Psychology, 51, 333-338. Clark, M. S., Ouellette, R., Powell, M. C., & Milberg, S. (1987). Recipient’s mood, relationship type and helping. Journal o f Personality and Social Psychology, 53, 94-103. Clark, M. S., & Reis, H. (1988). Interpersonal processes in close relationships. In M. R. Rosenzweig & L. Porter (Eds.), Annual review o f psychology (pp. 609-672). Palo Alto, CA: Annual Reviews. Clark, M. S., & Taraban, C. B. (1991). Reactions to and willingness to express emotion in communal and exchange relationships. Journal o f Experimental Social Psychology, 27, 324-336.

4.

COGNITIVE STRUCTURES AND PROCESSES

23 3

Clark, M. S., & Waddell, B. (1983). Effects of mood on thoughts about helping, attraction and in­ formation acquisition. Social Psychology Quarterly, 46, 31-35. Clark, M. S., & Waddell, B. (1985). Perceptions of exploitation in communal and exchange rela­ tionships. Journal o f Social and Personal Relationships, 2, 403-413. Clark, M. S., & Williamson, G. M. (1989). Moods and social judgements. In H. Wagner & A. Manstead (Eds.), Handbook o f social psychophysiology (pp. 347-370). New York: Wiley. Cohen, S., Sherrod, D. R., & Clark, M. S. (1986). Social skills and the stress-protective role of social support. Journal o f Personality and Social Psychology, 50, 963-973. Collins, N. L ., & Read, S. J. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal o f Personality and Social Psychology, 58, 644-663. Collins, N., & Read, S. (in press). Cognitive representations of attachment: The content and func­ tion of working models. In D. Perlman & K. Bartholomew (Eds.), Advances in personal relation­ ships: Vol. 5. Adult attachment relationships. London: Jessica Kingsley. Crandall, J. E. (1978). Effects of threat and failure on concern for others. Journal o f Research in Personality, 12, 350-360. Crocker, J., & Schwartz, I. (1985). Prejudice and ingroup favoritism in a minimal intergroup situa­ tion: Effects of self-esteem. Personality and Social Psychology Bulletin, 11, 379-386. Crocker, J., Thompson, L. L., McGraw, K. M ., & Ingerman, C. (1987). Downward comparison, prejudice, and evaluation of others: Effects of self-esteem and threat. Journal o f Personality and Social Psychology, 52, 907-916. Cunningham, M. (1988a). Does happiness mean friendliness? Induced mood and heterosexual self­ disclosure. Personality and Social Psychology Bulletin, 14, 283-297. Cunningham, M. (1988b). What do you do when you’re happy or blue: Mood, expectancies, and behavioral interest. Motivation and Emotion, 12, 309-330. Desmarais, S., & Lem er, M. J. (1989). A new look at equity and outcomes as determinants of satis­ faction in close personal relationships. Social Justice Research, 3, 105-119. Deutsch, M. (1975). Equity, equality and need: What determines which value will be used as the basis of distributive justice? Journal o f Social Issues, 31, 137-148. Deutsch, M. (1985). Distributive justice: A socio-psychological perspective. New Haven, CT: Yale University Press. Driscoll, R., Davis, K. E., & Lipitz, M. E. (1972). Parental interference and romantic love: The Romeo and Juliet effect. Journal o f Personality and Social Psychology, 24, 1-10. Dutton, D. G., & Aron, A. P. (1974). Some evidence for heightened sexual attraction under condi­ tions of high anxiety. Journal o f Personality and Social Psychology, 30, 510-517. Erber, R. (1991). Affective and semantic priming: Effects of mood category accessibility and infer­ ence. Journal o f Experimental Social Psychology, 27, 480-498. Erickson, M ., Sroufe, A., & Egeland, B. (1985). The relationship of quality of attachment and be­ havior problems in preschool children from a high risk sample. Monographs o f the Society fo r Research in Child Development, 205, 147-160. Feeney, J., & Noller, P. (1990). Attachment style as a predictor of adult romantic relationships. Journal o f Personality and Social Psychology, 58, 281-291. Festinger, L. (1951). Architecture and group membership. Journal o f Social Issues, 7, 152-163. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117-140. Fiedler, K., Pampe, H., & Scherf, U. (1986). Mood and memory for tightly organized social infor­ mation. European Journal o f Social Psychology, 16, 149-165. Fincham, F. D. (1985). Attribution processes in distressed and nondistressed couples: Vol. 2. Respon­ sibility for marital problems. Journal o f Abnormal Psychology, 94, 183-190. Fincham, F. D., & Beach, S. R. (1988). Attribution processes in distressed and nondistressed cou­ ples: Vol. 5. Real versus hypothetical events. Cognitive Therapy and Research, 12, 505-514. Fincham, F. D., Beach, S. R., & Baucom, D. H. (1987). Attribution processes in distressed and nondistressed couples: Vol. 4. Self-partner attribution differences. Journal o f Personality and So­ cial Psychology, 52, 739-748.

234

CLARK ET AL.

Fincham, F. D ., Beach, S. R., & Nelson, G. (1987). Attribution processes in distressed and nondis­ tressed couples: Vol. 3. Causal and responsibility attributions for spouse behavior. Cognitive Therapy and Research, 11, 71-86. Fincham, F. D., & Bradbury, T. N. (1988). The impact of attributions in marriage: An experimental analysis. Journal o f Social and Clinical Psychology, 7, 122-130. Fincham, F. D., & Bradbury, T. N. (1991). Cognition in marriage: A program of research on attri­ butions. Advances in Personal Relationships, 2, 159-203. Fincham, F. D., & Bradbury, T. N. (1992). Assessing attributions in marriage: The relationship attribution measure. Journal o f Personality and Social Psychology, 62, 457-468. Fincham, F. D., & O’Leary, K. D. (1983). Causal inferences for spouse behavior in maritally dis­ tress and nondistressed couples. Journal o f Social and Clinical Psychology, 1, 42-57. Fiske, A. P. (1992). The four elementary forms of sociality: Framework for a unified theory of so­ cial relations. Psychological Review, 99, 689-723. Fiske, A. P., Haslam, N ., & Fiske, S. T. (1991). Confusing one person with another: What errors reveal about the elementary forms of social relations. Journal o f Personality and Social Psycholo­ gy, 60, 656-674. Fletcher, G. O., & Fincham, F. D. (1991). Cognition in close relationships. Hillsdale, NJ: Lawrence Erlbaum Associates. Fonagy, P., Steele, H ., & Steele, M. (1991). Maternal representations of attachment during preg­ nancy predict the organization of infant-mother attachment at one year of age. Child Development, 62, 891-905. Forgas, J. P., & Bower, G. H. (1987). Mood effects on person perception judgments. Journal o f Personality and Social Psychology, 53, 53-60. Forgas, J. P ., Bower, G. H., & Krantz, S. E. (1984). The influence of mood on perceptions of social interactions. Journal o f Experimental Social Psychology, 20, 497-513. Forgas, J. P., & Moylan, S. (1987). After the movies: Transient mood and social judgment. Person­ ality and Social Psychology Bulletin, 4, 4 61-411.

Gaelick, L., Bodenhausen, G. V., & Wyer, R. S. (1985). Emotional communication in close rela­ tionships. Journal o f Personality and Social Psychology, 49, 1246-1265. George, C., Kaplan, N., & Main, M. (1985). The Berkeley Adult Attachment Interview. Unpublished protocol, Department of Psychology, University of California, Berkeley. Gottman, J. M ., & Levenson, R. W. (1988). The social psychophysiology of marriage. In P. Noller & M. A. Fitzpatrick (Eds.), Perspectives on m arital interaction (pp. 182-202). Cleveland, OH: Multilingual Matters. Greenberg, J. (1983). Equity and equality as clues to the relationship between exchange participants. European Journal o f Social Psychology, 13, 195-196. Griffitt, W. (1970). Environmental effects on interpersonal affective behavior: Ambient effective tem­ perature and attraction. Journal o f Personality and Social Psychology, 15, 240-244. Hammen, C. L., & Peters, S. D. (1977). Differential responses to male and female depressive reac­ tions. Journal o f Consulting and Clinical Psychology, 45, 994-1001. Hammen, C. L., & Peters, S. D. (1978). Interpersonal consequences of depression: Responses to men and women enacting a depressed role. Journal o f Abnormal Psychology, 87, 322-332. Hatfield, E., Traupmann, J., Sprecher, S., Utne, M ., & Hay, J. (1985). Equity and intimate rela­ tions: Recent research. In W. Ickes (Ed.), Compatible and incompatible relationships (pp. 91-117). New York: Springer-Verlag. Hatfield, E., Utne, M. K., & Traupmann, J. (1979). Social exchanges in developing relationships. In R. L. Burgess & T. L. Huston (Eds.), Equity theory and intimate relationships (pp. 99-135). New York: Academic Press. Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal o f Personality and Social Psychology, 52, 511-524. Heider, F. (1958). The psychology o f interpersonal relations. New York: Wiley. Helgeson, V. S. (1992). Moderators of the relation between perceived control and adjustment to chronic illness. Journal o f Personality and Social Psychology, 63, 656-666.

4.

COGNITIVE STRUCTURES AND PROCESSES

23 5

Hirt, E. R., Zillmann, D., Erickson, G. A., & Kennedy, C. (1992). Cost and benefits of allegiance: Changes in fan’s self-ascribed competencies after team victory versus defeat. Journal o f Personal­ ity and Social Psychology, 63, 724-738. Holmes, J. G. (1991). Trust and the appraisal process in close relationships. In W. H. Jones & D. Perl­ man (Eds.). Advances in personal relationships (pp. 57-107). London: Jessica Kingsley Publishers. Holmes, J. G., & Rempel, J. K. (1989). Trust in close relationships. In C. Hendrick (Ed.), Review o f personality and social psychology (pp. 187-220). Beverly Hills, CA: Sage. Isen, A. M. (1970). Success, failure, attention, and reaction to others. Journal o f Personality and Social Psychology, 15, 294-301. Isen, A. M. (1975). Positive affect, accessibility of cognitions and helping. In J. Piliavin (Chair), Current directions in theory on helping. Symposium presented at the meeting of the Eastern Psy­ chological Association, New York. Isen, A. M ., Clark, M. S., & Schwartz, M. F. (1976). Duration of the effect of good mood on help­ ing: Footprints on the sands of time. Journal o f Personality & Social Psychology, 36, 1-12. Isen, A. M., Shalker, T. E., Clark, M ., & Karp, L. (1978). Affect, accessibility of material in memory and behavior: A cognitive loop? Journal o f Personality and Social Psychology, 36, 1-12. Jacobson, N. S., McDonald, D. W ., Follette, W. C ., & Berley, R. A. (1985). Attributional process­ es in distressed and nondistressed married couples. Cognitive Therapy and Research, 9, 35-50. Johnson, D. J., & Rusbult, C. E. (1989). Resisting temptation: Devaluation of alternative partners as a means of maintaining commitment in close relationships. Journal o f Personality and Social Psychology, 57, 967-980. Kelley, H. H. (1979). Personal relationships: Their structures and processes. Hillsdale, NJ: Lawrence Erlbaum Associates. Kelley, H. H. (1984). Affect in relationships. In P. Shaver (Ed.), Review o f personality and social psychology: Emotions, relationships and health (Vol. 5, pp. 89-115). Beverly Hills, CA: Sage. Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J. H., Huston, T. L., Levinger, G., McClintock, E ., Peplau, L. A., & Peterson, D. R., (1983). Close relationships. New York: Freeman. Kelley, H. H ., & Thibaut, J. W. (1978). Interpersonal relationships: A theory o f interdependence. New York: Wiley. Knapp, A., & Clark, M. S. (1991). Some detrimental effects of negative mood on individuals’ ability to solve resource dilemmas. Personality and Social Psychology Bulletin, 17, 678-688. Lamb, M. E. (1987). Predictive implications of individual differences in attachment. Journal o f Con­ sulting and Clinical Psychology, 55, 817-824. Lamm, H., & Schwinger, T. (1980). Norms concerning distributive justice: Are needs taken into consideration in allocation decisions? Social Psychology Quarterly, 43, 425-429. Lamm, H., & Schwinger, T. (1983). Need consideration in allocation decisions: Is it just? Journal o f Social Psychology, 119, 205-209. Lerner, M. J., Miller, D. T ., & Holmes, J. G. (1976). Deserving and the emergence of forms of justice. In L. Berkowitz & E. Walster (Eds.), Advances in experimental social psychology (Vol. 9, pp. 133-162). New York: Academic Press. Lloyd, S., Cate, R. M ., & Henton, J. (1982). Equity and rewards as predictors of satisfaction in casual and intimate relationships. The Journal o f Psychology, 110, 43-48. Lujansky, H., & Mikula, G. (1983). Can equity explain the quality and stability of romantic relation­ ships? British Journal o f Social Psychology, 22, 101-112. Mehta, P., Clark, M. S., & Mills, J. (1992). Interest in time another spends making a selection fo r oneself: The impact o f relationship type. Unpublished manuscript. Mikulincer, M ., & Erev, I. (1991). Attachment styles and the structure of romantic love. British Journal o f Social Psychology, 30, 273-291. Mills, J., & Clark, M. S. (1982). Communal and exchange relationships. In L. Wheeler (Ed.), Review o f personality and social psychology (pp. 121-144). Beverly Hills, CA: Sage. Mills, J., Clark, M. S., & Ford, T. (1992). Developm ent o f a measure o f strength o f communal rela­ tionships. Unpublished manuscript.

23 6

CLARK ET AL.

Murray, S. L., & Holmes, J. G. (1992). Seeing virtues in faults: N egativity and the transformation o f interpersonal narratives in close relationships. Unpublished manuscript, University of Water­ loo, Waterloo, Ontario. Noller, P. (1980). Misunderstandings in marital communication: A study of couples’ nonverbal com­ munication. Journal o f Personality and Social Psychology, 39, 1135-1148. Noller, P., & Ruzzene, M. (1991). Communication in marriage: The influence of affect and cogni­ tion. In G. Fletcher & F. D. Fincham (Eds.), Cognition in close relationships (pp. 203-233). Hills­ dale, NJ: Lawrence Erlbaum Associates. Pleban, R., & Tesser, A. (1981). The effects of relevance and quality of another’s performance on interpersonal closeness. Social Psychology Quarterly, 44, 278-285. Procidano, M. E ., & Heller, K. (1983). Measures of perceived social support from friends and from family: Three validation studies. American Journal o f Community Psychology, 11, 1-24. Reis, H. T. (1990). The role of intimacy in interpersonal relations. Journal o f Social and Clinical Psychology, 9, 15-30. Reis, H. T ., & Shaver, P. R. (1988). Intimacy as an interpersonal process. In S. W. Duck (Ed.), Handbook o f personal relationships (pp. 367-389). New York: Wiley. Rosen, L. D., Fiske, S., & Rosenthal, A. (1991, August). Romantic outcome dependency and the (in)accuracy o f impression form ation: A case o f clouded judgm ent. Poster presented at American Psychological Association meeting, Boston, Massachusetts. Ruble, D. N., & Thompson, E. P. (1992). The implications of research on social development for mental health: An internal socialization perspective. In D. N. Ruble, P. R. Costanzo, & M. E. Oliveri (Eds.), The social psychology o f mental health: Basic mechanisms and applications (pp. 81-125). New York: Guilford. Rusbult, C. E. (1980a). Commitment and satisfaction in romantic associations: A test of the invest­ ment model. Journal o f Experimental Social Psychology, 16, 172-186. Rusbult, C. E. (1980b). Satisfaction and commitment in friendships. Representative Research in So­ cial Psychology, 11, 96-105. Rusbult, C. E. (1983). A longitudinal test of the investment model: The development (and deteriora­ tion) of satisfaction and commitment in heterosexual involvements. Journal o f Personality and So­ cial Psychology, 45, 101-117. Rusbult, C. E., & Buunk, B. P. (in press). Commitment processes in close relationships: An inter­ dependence analysis. Journal o f Social and Personal Relationships. Rusbult, C. E., & Farrell, D. (1983). A longitudinal test of the investment model: The impact of job satisfaction, job commitment, and turnover of variations in rewards, costs, alternatives, and investments. Journal o f A pplied Psychology, 68, 429-438. Rusbult, C. E., Johnson, D. J., & Morrow, G. D. (1986). Predicting satisfaction and commitment in adult romantic involvements: An assessment of the generalizability of the investment model. Social Psychology Quarterly, 49, 81-89. Rusbult, C. E., & Martz, J. (1992). The decision to remain in an abusive relationship: An investment m odel analysis. Unpublished manuscript, University of North Carolina at Chapel Hill. Rusbult, C. E., Verette, J., Whitney, G. A., Slovik, L. F., & Lipkus, I. (1991). Accommodation processes in close relationships: Theory and preliminary empirical evidence. Journal o f Personal­ ity and Social Psychology, 60, 53-78. Sabatelli, R. M ., & Cecil-Pigo, E. F. (1985). Relational interdependence and commitment in mar­ riage. Journal o f M arriage and the Family, 47, 931-937. Schachter, S. (1964). The interaction of cognitive and physiological determinants of emotional state. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 49-80). New York: Aca­ demic Press. Schachter, S., & Singer, J. F. (1962). Cognitive, social and physiological determinants of emotional state. Psychological Review, 69, 379-399. Schafer, R. B., & Keith, P. M. (1981). Equity in marital roles across the family cycle. Journal o f M arriage and the Family, 43, 359-367.

4.

COGNITIVE STRUCTURES AND PROCESSES

237

Schank, R., & Abelson, R. (1977). Scripts, plans and knowledge. Hillsdale, NJ: Lawrence Erlbaum Associates. Schiffenbauer, A. (1974). Effect of observer’s emotional state on judgments of the emotional states of others. Journal o f Personality and Social Psychology, 30, 31-35. Schwartz, J. C., & Shaver, P. (1987). Emotions and emotion knowledge in interpersonal relations. In W. Jones & D. Perlman (Eds.), Advances in personal relationships (pp. 197-241). Greenwich, CT: JAI. Schwarz, N., & Clore, G. L. (1983). Mood, misattribution and judgments of well-being: Informa­ tive and directive functions of affective states. Journal o f Personality and Social Psychology, 45, 513-523. Schwarz, N., & Clore, G. L. (1987). How do I feel about it? The informative function of affective states. In K. Fielder & J. Forgas (Eds.), Affect, cognition and social behavior: New evidence and integrative attempts (pp. 44-61). Toronto: C. J. Hogrefe. Segal, M. W. (1974). Alphabet and attraction: An unobtrusive measure of the effect of propinquity in a field setting. Journal o f Personality and Social Psychology, 30, 654-657. Seligman, C., Fazio, R. H ., & Zanna, M. P. (1980). Effects of salience of extrinsic rewards on liking and loving. Journal o f Personality and Social Psychology, 38, 453-460. Shaver, K. G. (1985). The attribution o f blame: Causality, responsibility, and blameworthiness. New York: Springer-Verlag. Shaver, P. R., & Hazan, C. (1992). Adult romantic attachment: Theory and evidence. In D. Perlman & W. Jones (Eds.), Advances in personal relationships (Vol. 4, pp. 29-70), Greenwich, CT: JAI. Shaver, P. R., Hazan, C., & Bradshaw, D. (1988). Love as attachment: The integration of three behavioral systems. In R. J. Sternberg & M. L. Barnes (Eds.), The psychology o f love (pp. 68-99). New Haven, CT: Yale University Press. Shaver, P., Schwartz, J., Kirson, D., & O’Connor, C. (1987). Emotion knowledge: Further explo­ ration of a prototype approach. Journal o f Personality and Social Psychology, 52, 1061-1086. Simpson, J. A. (1987). The dissolution of romantic relationships: Factors involved in relationship stability and emotional distress. Journal o f Personality and Social Psychology, 39, 683-692. Simpson, J. A., Gangestad, S. W ., & Lerma, M. (1990). Perception of physical attractiveness: Mecha­ nisms involved in the maintenance of romantic relationships. Journal o f Personality and Social Psychology, 59, 1192-1201. Snyder, M ., Tanke, E. D., & Berscheid, E. (1977). Social perception and interpersonal behavior: On the self-fulfilling nature of social stereotypes. Journal o f Personality and Social Psychology, 35, 656-666. Sommers, S. (1984). Reported emotions and conventions of emotionality among college students. Journal o f Personality and Social Psychology, 46, 207-215. Sprecher, S. (1986). The relation between inequity and emotions in close relationships. Social Psy­ chology Quarterly, 49, 309-321. Sprecher, S. (1988). Investment model, equity and social support determinants of relationship com­ mitment. Social Psychology Quarterly, 51, 318-328. Taylor, S. E. (1983). Adjustments to threatening events: A theory of cognitive adaption. American Psychologist, 38, 1161-1173. Taylor, S. E., Kemeny, M. E ., Aspinwall, L. G., Schneider, S. G., Rodriguez, R., & Herbert, M. (1992). Optimism, coping, psychological distress, and high-risk sexual behavior among men at risk for Acquired Immunodeficiency Syndrome (AIDS). Journal o f Personality and Social P sy­ chology, 63, 460-473. Taylor, S. E., & Lobel, M. (1989). Social comparison activity under threat: Downward evaluation and upward contacts. Psychological Review, 96, 569-575. Tesser, A. (1980). Self-esteem maintenance in family dynamics. Journal o f Personality and Social Psychology, 39, 77-91. Tesser, A. (1988). Toward a self examination maintenance model of social behavior. Advances in Experimental Social Psychology, 21, 181-227.

238

CLARK ET AL.

Tesser, A., & Campbell, J. (1980). Self-definition: The impact of the relative performance and similarity of others. Social Psychology Quarterly, 43, 341-347. Tesser, A., & Campbell, J. (1982). Self-evaluation maintenance and the perception of friends and strangers. Journal o f Personality, 59, 261-279. Tesser, A ., & Collins, J. E. (1988). Emotion in social reflection and comparison situations: Intui­ tive, systematic and exploratory approaches. Journal o f Personality and Social Psychology, 55, 695-709. Tesser, A., & Cornell, D. P. (1991). On the confluences of self-processes. Journal o f Experimental Social Psychology, 27, 501-526. Tesser, A., & Cowan, C. L. (1975). Some effects of thought and number of cognitions on attitude change. Social Behaviors and Personality, 31, 165-173. Tesser, A., Millar, M ., & Moore, J. (1988). Some affective consequences of social comparison and reflection processes: The pain and pleasure of being close. Journal o f Personality and Social P sy­ chology, 54, 49-61. Tesser, A., & Paulhus, D. L. (1983). The definition of self: Private and public self-evaluation manage­ ment strategies. Journal o f Personality and Social Psychology, 44, 672-682. Tesser, A., Pilkington, C. J., & McIntosh, W. D. (1989). Self-evaluation maintenance and the mediational role of emotion: The perception of friends and strangers. Journal o f Personality and Social Psychology, 57, 442-456. Tesser, A., & Smith, J. (1980). Some effects of friendship and task relevance on helping: You don’t always help the one you like. Journal o f Experimental Social Psychology, 16, 582-590. Thibaut, J. W ., & Kelley, H. H. (1959). The social psychology o f groups. New York: Wiley. Thompson, S. C., & Kelley, H. H. (1981). Judgments of responsibility for activities in close rela­ tionships. Journal o f Personality and Social Psychology, 41, 469-477. Traupmann, J., Hatfield, E., & Sprecher, S. (1981). The importance o f fairn ess”fo r marital satis­ faction o f older women. Unpublished manuscript. Utne, M. K., Hatfield, E ., Traupmann, J., & Greenberger, D. (1984). Equity, marital satisfaction, and stability. Journal o f Social and Personal Relationships, 1, 323-332. Walster, E., & Berscheid, E. (1971). Adrenaline makes the heart grow fonder. Psychology Today, 5, 47-62. Walster, E ., Walster, G. W ., & Berscheid, E. (1978). Equity: Theory and research. Boston: Allyn & Bacon. Waters, E., Wippman, J., & Sroufe, L. A. (1979). Attachment, positive affect, and competence in the peer group: Two studies in construct validation. Child Development, 50, 821-829. Weiner, B., Freize, I. H., Kukla, A., Reed, L ., Rest, S., & Rosenbaum, R. M. (1972). Perceiving the causes of success and failure. In E. E. Jones, D. E. Kanouse, H. H. Kelley, R. E. Nisbett, S. Valins, & B. Weiner (Eds.), N ew directions in attribution research (Vol. 2, pp. 95-120). Mor­ ristown, NJ: General Learning Press. White, G. L., Fishbein, S., & Rutstein, J. (1981). Passionate love and the misattribution of arousal. Journal o f Personality and Social Psychology, 41, 56-62. Williamson, G. M ., & Clark, M. S. (1989). Providing help and desired relationship type as deter­ minants of changes in moods and self-evaluations. Journal o f Personality and Social Psychology, 56, 722-734. Williamson, G. M ., & Clark, M. S. (1992). Impact of desired relationship type on affective reac­ tions to choosing and being required to help. Personality and Social Psychology Bulletin, 18, 10-18. Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90, 245-271. Wood, J. V. (1989). Theory and research concerning social comparisons of personal attributes. Psy­ chological Bulletin, 106, 231-248.

5

Social Cognition and Health Psychology

Leslie F. Clark Purdue University

Contents Shared Features of Social Cognition and Health Psychology 240 . Contents of This Chapter 240 Attentional Processes in Health Psychology 242 Interpretations and Reactions to Somatic States 242 The Role of Attention in Long-Term Adjustment Benefits 243 The Consequences of Receiving Illness Labels 248 Minimization and Confirmatory Search Processes 249 Framing Biases and Other Contextual Factors Influencing Interpretation 250 Mental Representations Relevant to Health and Illness 251 Representations of Sensations and Symptoms 251 Representations of Illnesses 252 Representations of Individuals Experiencing an Illness 255 Representations of the Self and Implications for Illness 257 Capturing Individuals’ Mental Models of Health Risks 259 The Role of Affect in Health Cognition, Memory, and Adjustment 260 The Role of Affect in Health-Relevant Judgments and Memory 261 Stable Affect-Relevant Orientations Toward Information Processing 263 Social Comparison and Health Psychology 267 Symptom Appraisal 268 Reactions to Diagnosis Information 268 Recovery and Adjustment to Illness 270 Summary 274 Potential Theoretical Integrations in Two Domains in Health Psychology 274 Health Psychology’s Contributions to Social Cognition 278 Conclusions 279 Acknowledgments 279 References 279

This chapter addresses work in health psychology from a social cognition per­ spective. Health psychology has been defined as the field in which “psychology

239

240

CLARK

is devoted to understanding psychological influences on how people stay healthy, why they become ill, and how they respond when they do get ill” (Taylor, 1991, p. 6). Some researchers in health psychology identify their work as “social cog­ nition,” whereas others may not consciously do so. Indeed, health psychology has been influenced in many ways by the same concerns and research in cogni­ tive psychology that have influenced work in the area of social cognition (Fiske & Taylor, 1991; Martin & Clark, 1990). Much of the work in health psychology has developed simultaneously with the cognitively influenced strides that have been made in other areas of social psychology. This chapter represents perhaps the first attempt to cover the topic of social cognition in health psychology. I discuss work in health psychology that clearly derives from earlier work in “so­ cial cognition,” as well as areas of health psychology that share similar concerns and theoretical perspectives.

SHARED FEATURES OF SOCIAL COGNITION AND HEALTH PSYCHOLOGY Social cognition and health psychology research share three common features. First, each field has a predisposition toward intraindividualistic explanations (i.e., a focus on the internal, psychological processes of the individual). Individuals are seen as constructing their own perceptions of reality and then operating with­ in that constructed reality. More specifically, important roles are played by per­ ceptual, interpretational, inferential, memorial, judgmental, and decision making processes in the theories of both areas. A second area of overlap concerns the type of theories that each field favors. Both areas are interested in the development of explanations that detail the process­ es described previously (see Martin & Clark, 1990; Taylor, 1991) and in theo­ retical frameworks that can accommodate the rich interplay among these processes. Third, both areas harbor the overall goal of understanding the relationships among cognition, affect, and behavior. In the health area, this is easily seen in work on the labeling of symptoms and subsequent decisions to seek medical care and in the work on adherence to medical regimens.

CONTENTS OF THIS CHAPTER This chapter discusses research using assumptions made by information process­ ing frameworks (see Wyer & Srull, 1989). Like most information processing frameworks, this one suggests that processing objectives determine how stimuli are interpreted and encoded. In addition, stimuli are interpreted and encoded in terms of the perceivers’ preexisting knowledge, which is stored in the form of memory representations. Interpretational processes lead to an encoding of the

5.

241

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

stimulus information into a memory representation that is stored in long-term memory. Judgments such as "this is threatening information" or "I am probably sick," which are based on this encoding and one's relevant preexisting knowledge, determine the individual's emotional, behavioral, and cognitive responses to the stimulus. Health psychology research emphasizes additional components to such a framework. These have been included in a diagram in Fig. 5.1. Boxes denote states, constructs, or outcomes, and octagons denote processes. First, antecedent factors in the form of personality distinctions, such as optimism, or situational factors, such as mood or anxiety, playa significant role in information processing. These antecedent factors help to set one's processing goals and, in addition, directly affect the way information is interpreted. For example, an antecedent confronting coping style may lead to a processing goal of "attend carefully to what the physician is saying about my diagnosis." At the same time, one's anxiety level may interfere directly with the individual's ability to comprehend what the physician is saying. These antecedent factors also may influence memory search in terms of the confirmatory search strategies and mood-congruent memory effects that are discussed in detail later in this chapter. Second, health research emphasizes the surrounding context in which stimulus information is presented. For example, the affect of the physician as he or she is delivering diagnosis information can influence how that information is processed. This chapter provides examples of this in terms of framing effects and social comparison information. Third, health research dramatically reminds us of how one's responses can affect future information processing. For example, the emotional, cognitive, and behavioral responses to the diagnosis of cancer

Retrieve and Search Memory Processing

Representations

Objectives

Antecedent Person end

o 0

Factors

c

Process Information (e.g. encodng,

Mental Representation

3

Compute Judgement

Information

Surroundng Context

FIG. 5.1.

(0

ofStim iJus

interpretation)

SHmulue

1 3)

An information processing framework for health-relevant stimuli.

i

Behavioral, Emotional, and

Situational

242

CLARK

will undoubtedly set the stage for antecedent factors that will affect the process­ ing of information about treatment decisions, comparison others, and other health threats. These additions were inspired by Lazarus and Folkman’s (1984) model of stress appraisal processes. Each area of research to be discussed is linked to this overall framework. Health psychology is a very large interdisciplinary area. Therefore, some boundaries must be drawn for the purposes of this chapter. I discuss work from the perspective of a single individual who might be either a medical patient or a healthy person. Five topic major areas have been selected for discussion. These areas are by no means exhaustive of the work that integrates social cognition with health and illness concerns. However, the areas represent ongoing research that exhibits a distinctly social cognition flavor. The first two areas include attentional and interpretational processes. The third area examines various mental represen­ tations that are relevant to health and illness. The fourth area focuses on the influence of affect on memory, judgment, and health-relevant cognition. This sec­ tion covers research concerning the effects of mood on health-related memory, affective orientations toward information processing about disease, and the role that these orientations play in adaptation to illness. The fifth area highlights work derived from social comparison theory, which constitutes a vast body of research on cognitive processes that has developed over decades prior to the advent of social cognition per se. However, research on social comparisons has shown a recent resurgence in the context of health psychology. Its concerns, dovetail those of a social cognition approach, and therefore are worthy of attention. None of these sections provides an exhaustive review. Instead, each presents current representative empirical work, along with a discussion of relevant the­ ories and mechanisms. ATTENTIONAL PROCESSES IN HEALTH PSYCHOLOGY Attentional processes, including distraction, avoidance, and minimization, have been widely studied in several areas of health psychology. One may think of at­ tention in terms of one’s predisposition to attend to information in threatening situations (an antecedent factor) or in terms of stimulus characteristics that pro­ voke specific attention, such as the provision of information about the sensations a surgical patient might be expected to experience (Suls & Fletcher, 1985; Tay­ lor & Clark, 1986). The three domains discussed herein include: (a) the interpre­ tation of somatic states, (b) adjustment to long-term health problems, and (c) reactions to health-relevant diagnosis labels. Interpretations and Reactions to Somatic States

The interpretation of somatic states is extremely important in health psychology. Individuals use their interpretations of somatic states as input to their conceptu­ alizations of chronic diseases and to inform themselves about their own health

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

24 3

status (Cioffi, 1991a; Skelton & Pennebaker, 1990). Therefore, these interpreta­ tions guide health-relevant behaviors involved in self-care or seeking profession­ al health care. For example, diabetic individuals tend to rely more on their own perception of symptoms than on actual objective measures of blood glucose to monitor their diabetes (Gonder-Frederick, Cox, Bobbitt, & Pennebaker, 1986). Empirical research on the interpretation of somatic states has accrued from several sources. First, researchers using laboratory pain tasks have examined var­ ious strategies that increase tolerance for pain and decrease perceived discomfort (Leventhal & Johnson, 1983). In addition, field research in health psychology has examined the perception of pain and discomfort in settings involving a wide variety of medical procedures, including surgery (Johnson, 1975; Suls & Fletch­ er, 1985; Taylor & Clark, 1986). Attention Versus Avoidance. In the laboratory, research on pain tolerance requiring individuals to hold their hands in extremely cold water demonstrated that attention to concrete sensation information increases people’s pain tolerance and reduces distress during this cold pressor task (Johnson & Leventhal, 1974; Leventhal, Brown, Shacham, & Engquist, 1979). Field studies of clinical dis­ tress also have suggested that attention focused on sensations often decreases dis­ tress more than focusing on distracting images or emotional responses (see Cioffi, 1991a; Johnson, 1984; Taylor & Clark, 1986 for reviews). In a review of studies examining the provision of information about sensa­ tions to be experienced in medical settings, Taylor and Clark (1986) found that information given to patients about sensations, the medical procedure, or both was more helpful than no information or emotional support. In addition, Taylor and Clark reported that sensation information consistently improved adaptation on measures such as ambulation after surgery, amount of postoperative pain medi­ cation needed, and self-reports of mood. Physiological indicators of distress (e.g., heart rate) and self-reports of pain or discomfort were less reliably affected. They concluded that sensation information alone was more effective than information about the medical procedure (see also Johnson, 1984). In a meta-analysis that reviewed dozens of lab and field studies, Suls and Fletch­ er (1985) concluded that when individuals attended to their emotional reactions or were given no explicit attentional set, attention clearly led to worse outcomes than distraction techniques. However, when individuals focused on the sensations as­ sociated with the noxious procedure/event, they fared better than those in distrac­ tion groups. The consensus from both laboratory and field research is that actively focusing on the sensations that one experiences will, in general, be beneficial. The Role of Attention in Long-Term Adjustment Benefits

In terms of long-term adjustment, avoidance has often been operationalized as denial: the lack of recognition that one is experiencing on illness condition, or that the condition is severe in its health implications. Some studies have found

244

CLARK

that such denial is helpful (Meyerowitz, 1983), whereas others have found that it has negative effects. For example, Namir, Wolcott, Fawzy, and Alumbaugh (1987) studied the impact of denial on adjustment to human immunovirus (HIV) infection. They found that deniers did not experience less emotional distress or fewer auto immune deficiency syndrome (AIDS)-related worries than nondeniers. Deniers showed a greater tendency to experience depression and self-esteem problems than nondeniers. In contrast, an active behavioral coping style was posi­ tively associated with mood benefits and greater self-esteem. On the other hand, denial also has been found to be helpful. In a study of cancer patients, Meyerowitz (1983) operationalized cancer-specific denial as a coping strategy in which individuals avoid stressful information about cancer and minimize their emotional reactions to the cancer. Individuals who exhibited this specific sense of denial reported lower emotional distress and physical discomfort than did individuals who did not exhibit this denial. In addition, she found that the muted emotional upset when thinking about the cancer was responsible for the coping benefits and not the amount of thinking about the cancer per se. Measures of one’s general tendency to deny stress did not show these effects. Similar benefits for specific denial have been found for heart attack patients (Krantz & Deckel, 1983). The pattern for surgery studies showed that denial may have some initial benefits in reducing anxiety, but that attending to the stress of surgery shortens one’s postoperative hospital stay (Johnson, 1984; Johnson, Lauver, & Nail, 1989). In their meta-analysis, Suls and Fletcher (1985) reported that long-term outcomes were more favorable with manipulations or personality predispositions that en­ hanced attention than they were with those that enhanced distraction as a coping response. This suggests that avoidance, distraction, or repressive coping styles decline in their effectiveness over time (see also Mullen & Suls, 1982). These conclusions match those found with myocardial infarction patients by Levine et al. (1988). Individuals high in denial initially showed benefits in the form of shortened stays in intensive care and fewer signs of cardiac dysfunction. However, years later, these same individuals complied less with their medical regimen and spent more time in the hospital than individuals who were not high on denial. Overall, research on denial has indicated that it may interfere with long-term adaptation by disrupting: (a) symptom monitoring, (b) compliance to medical treatment, and (c) attempts to make major lifestyle changes (Garrity et al., 1976; Taylor, 1991). Theoretical History for Attentional Benefits. Several theoretical accounts for why attention produces benefits have been offered (see Cioffi, 1991a for a review). For example, Janis (1958) offered a “work of worry” hypothesis to explain the superior adjustment of moderately fearful surgery patients compared with those who were very low in presurgery fears. This finding was interpreted in terms of the mental preparation that these patients engaged in prior to the

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

24 5

surgery. Specifically, the surgery stimulated worry in the patients, motivating them to mentally prepare for the side effects and difficulties associated with the noxious event. The exact cognitive processes involved in this preparation never were out­ lined explicitly. Moreover, such an explanation seems less plausible in account­ ing for the beneficial effects found in cold pressor laboratory tasks, where individuals do not need to do anything to complete the task (Cioffi, 1993). In addition, these surgical patients were not systematically prepared for the sensa­ tions that they would be experiencing. Researchers have concluded that sensa­ tion information, and not simply information about the aversive procedure, creates the adaptive benefits (Johnson, 1984; Suls & Fletcher, 1985; Suls & Wan, 1982; Taylor & Clark, 1986). Even so, aspects of Janis’ early explanation have found their way into more recent accounts of why attention is beneficial. Second, the construct of psychological control has been offered as an expla­ nation of why providing information that stimulates attention is adaptive (see Thompson, 1981, for a review). In particular, different forms of control have been identified as possible mechanisms. It has been suggested that providing in­ formation may help one experience less reactance or helplessness in a stressful situation. However, measures of anger and helplessness in medical setting studies do not reliably show reductions when individuals are provided with information, failing to provide support for this version of psychological control (Taylor & Clark, 1986). Overall, there is little direct evidence that information manipulations in fact enhance feelings of control (Cioffi, 1993). There is an additional version of the psychological control explanation in which information increases the individual’s ability to predict what will happen (or the reactions to it). The creation of this mental model of what to expect decreases the individual’s subsequent surprise or arousal (Miller, 1979). This is similar to Johnson’s (1975) notion of a “congruence” hypothesis, in which information can reduce the discrepancy between expected and experienced sensations. Contrary to the congruence hypothesis, Leventhal et al. (1979) found that pain warnings, which might be expected to decrease surprise and distress by generating accurate expectations, instead blocked the normally beneficial effects of sensation infor­ mation. Thompson (1981) also suggested that such an explanation of predictabil­ ity cannot account for reductions in anticipatory distress, because the event has not happened yet. Leventhal (1982) outlined a “parallel process” model in which individuals process information about the objective qualities of a sensation at the same time as they perceive it along dimensions of emotional distress and threat value. Focus­ ing individuals on objective sensation information serves to decrease the stimu­ lus’ potential for generating emotional distress. In essence, the individual forms a mental schema in which the informational properties of the stimuli are encoded while blocking the encoding of emotional information. Therefore, the overall event is experienced as less distressing. Like the Miller (1979) and Johnson (1984)

246

CLARK

hypotheses, Leventhal also predicted benefits in terms of reduced surprise and arousal. The last two explanations rely on the concept of schema or mental representa­ tions regarding one’s experience of the stimulus, much like work in schemas in social cognition (Brewer & Nakamura, 1948; Fiske & Taylor, 1991; Hastie, 1981). This shows the increasing influence of social cognition and cognitive psychology on the research regarding one’s interpretations of somatic states. Two recent for­ mulations described next show this influence to an even greater extent. They also represent a call to researchers to capture the complex and interactive influences of such top-down processes along with the bottom-up influence of surrounding contextual factors in ongoing research. Recent Theoretical Integrations. One of these theoretical integrations is couched in terms of control systems theory (Suls & Fletcher, 1985), and has been referred to as the “self-regulation model” (Leventhal & Johnson, 1983). This ap­ proach argues that cognitive processes operate to affect several outcomes. The self-regulation model incorporates the parallel process model described earlier, in which individuals provided with sensation information are more capable of developing a schema of the experience based on its objective properties and not its distressing emotional implications. In addition, this schema provides useful information regarding efficacious coping options (see Johnson, 1984; Suls & Fletcher, 1985). Specifically, a negative feedback loop enables individuals to per­ ceive symptoms in terms of a discrepancy between a standard of ongoing good health and one’s present state. This discrepancy motivates coping efforts based on the relevant information at hand. Finally, the point is made that it may take time for the schema representing objective features of the stimuli to override long­ standing emotional reactions/interpretations to such stimuli (Leventhal, 1982). Suls and Fletcher (1985) suggested that this is the reason that attentional strate­ gies show the greatest adaptational effects on long-term, rather than short-term measures. In a second theoretical integration, Cioffi (1991a) described a cognitiveperceptual model of somatic interpretation. This comprehensive model took into account the situations under which somatic information will be salient to an in­ dividual, as well as the factors that determine interpretation and labeling of the state as a symptom. Individuals mentally elaborate an experienced physiological state. This elaboration is affected by one’s coping strategies, activated goals, and represented knowledge about illness such as health hypotheses and attributions regarding causes and consequences. These top-down processes help determine what aspects of the sensation will be attended to, as well as how they will be interpreted. At the same time, actual stimulus properties, such as its persistence or the degree to which it disrupts one’s activities, also affect its interpretation. Finally, the surrounding context in which the sensation is experienced will make salient

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

24 7

certain aspects of the sensation, thus influencing how it is finally categorized (e.g., as an illness symptom, or as a bothersome, but not dangerous, situation). Cioffi made the point that some situations are overdetermined (i.e., there are few alter­ native explanations for the causes of one’s sensations). An example of such a sit­ uation is the laboratory experiment using the cold pressor task, in which reasoning about the sources and causes of one’s painful sensation is determined easily. Over­ all, Cioffi’s model suggests that the content of one’s attention has been overlooked in past research and constitutes a critical aspect of one’s response to physiologi­ cal states. Additional Directions. In her review of work on attention and avoidance, Cioffi (1993) offered several ideas that are only beginning to be researched and provide intriguing future directions for researchers of somatic experience. The first of these draws on the gate control theory of pain (Melzack & Wall, 1982). Attending to a primary sensation may also make one aware of the whole family of sensations in the painful area. This awareness may diminish the experienced pain via gate-control principles. Specifically, stimulating a neurologically com­ peting pathway will inhibit signals from other sites, including the primary pain site. Thus, attending to secondary sensations may cause a weaker neural signal to leave the primary stimulus site. Therefore, the actual experience of the painful sensation is lessened. The second idea draws on action identification theory, developed by Vallacher and Wegner (1985, 1987), in which actions and experiences can be conceptual­ ized at either a low concrete level of meaning or at higher levels. For example, the experience of a somatic state may be conceptualized as either a “sense of pres­ sure against the inside of one’s eye socket” (low level identification) or as a “stressinduced headache” (higher level identification). Cioffi (1991a, 1993) has suggested that sensory monitoring, or attending to one’s sensations, may be effective, because it first acquaints individuals with lower level, concrete experience. After such familiarity, individuals can more easily make a benign higher-order interpretation of the experience rather than the more threatening (default) interpretation that they initially held. For example, for some, being in a doctor’s presence where one puts faith in his or her knowledge and abilities may lead to a reinterpretation of threatening aspects of a sensation. Ac­ tion Identification Theory predicts that such a replacement of higher order in­ terpretations is most likely when one has first identified the experience in lower level concrete terms. According to Cioffi, the provision of sensation information may serve this function. Finally, Cioffi and Holloway (1993) have applied the findings of Wegner (1992) regarding thought suppression to help explain why distraction techniques are doomed to fail. Specifically, Cioffi and Holloway used a cold pressor task in which they instructed subjects either to (a) imagine in detail their rooms at home, (b) pay close attention to the sensations in their hands, or (c) avoid attending to the

24 8

CLARK

sensations in their hands. These instructions represented the distraction, sensa­ tion monitoring, and suppression conditions, respectively. Subjects in the monitor­ ing condition showed the fastest pain recovery during the 2 minutes after the cold pressor task, whereas the suppression condition produced the slowest recovery. In addition, the suppression subjects were more likely than other subjects to in­ terpret a subsequent innocuous vibration as unpleasant. These findings suggest that forcibly trying to remove a thought about a sensa­ tion leads to a “rebound effect,” in which thoughts about the distressing sensation enter consciousness more often than they would in either a distraction task or monitoring task (Wegner, 1992). If, during an attempt at distraction or avoidance, individuals find themselves thinking about the stimuli and therefore try to remove such thoughts, rebound effects may lead them to be worse off later than if they originally deliberately focused on the stimuli. This mechanism may account for the finding reported by Namir et al. (1987) that individuals who attempt to main­ tain a positive mind set by not thinking of negative aspects of AIDS reported experiencing more obsessive thoughts about the disease than did individuals who did not use this coping strategy. The next section examines work on reactions to illness diagnosis labels.

THE CONSEQUENCES OF RECEIVING ILLNESS LABELS When individuals are labeled with a particular health diagnosis, it affects both their behaviors and cognitions. For example, actually experiencing an illness less­ ens one’s perception of its seriousness (Jemmott, Croyle, & Ditto, 1988). Perhaps actual experience includes the knowledge that one has survived the illness and therefore makes the condition seem less serious than an unexperienced illness. However, it is also possible that individuals attempt to downplay experiences that they have found (or anticipate to be) threatening. In either case, these memory representations of past experiences appear to affect the way in which individuals perceive current labeling information (see Fig. 5.1). Researchers in the area of coping have suggested that minimization or avoidance of the seriousness of a problem can play a role in reducing arousal and distress to a more manageable level where it does not interfere with treatment and other coping actions (Cohen & Lazarus, 1973; Lazarus, 1983). Research in the labora­ tory has used false feedback regarding health diagnoses to examine the reactions of individuals to such information. These studies manipulated the diagnosis of a fictitious enzyme deficiency (Croyle & Sande, 1988; Ditto & Jemmott, 1989; Ditto, Jemmott, & Darley, 1988; Jemmott, Ditto, & Croyle, 1986), high blood pressure (Croyle, 1990), and cholesterol levels (Croyle, Sun, & Louie, in press). These studies all showed that individuals who believe themselves to be at risk down­ play the seriousness of the disorder, while at the same time showing evidence of a confirmatory search strategy for diagnosis-relevant symptoms or risk behaviors.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

24 9

Minimization and Confirmatory Search Processes

Evidence for individuals’ minimization of a health threat has been found across several measures (Croyle, 1990; Croyle & Ditto, 1990; Croyle & Jemmott, 1991; Croyle & Sande, 1988). Individuals who believed they tested positive for a nega­ tive health condition rated that condition as being less serious and tended to deni­ grate the test’s validity more than those who believed that they tested negative (Cioffi, 1991b; Croyle & Sande, 1988). However, these two groups of subjects showed no differences in actual affect. This null effect has been interpreted as a blunting of affect for individuals who tested positive. Individuals who down­ played the seriousness of the condition also had a tendency to see the condition as being cyclic, rather than chronic (Croyle, 1990). Finally, subjects distorted the recall of their positive test results in a more negative direction (Ditto et al., 1988). Croyle and Sande (1988) manipulated the reliability of the test results by tell­ ing subjects it was either 95% or 75% accurate. Test reliability did not affect seriousness ratings, but it did affect information-seeking intentions. Subjects who tested positive were less likely to request a definitive follow-up when the test was relatively unreliable. Individuals who were led to believe that the test might be wrong were less anxious to discover their true condition by taking a more definitive follow-up test (although see Cioffi, 1991b). Perhaps the most intriguing demonstration of the tendency to minimize the seriousness of positive test results was offered by Ditto et al. (1988). Some sub­ jects were told that an easy treatment existed for handling an obscure enzyme deficiency, whereas the remaining subjects were told nothing about treatment. Subjects given the treatment information rated the condition as more serious when they believed they had the deficiency than when they did not. On the other hand, for subjects who were not given the treatment information, those who believed they had the condition rated the condition as less serious than those who believed they had tested negative for the condition. These results indicate that minimiza­ tion is occurring normally; the knowledge of treatment lowered the perceived health threat and hence individuals’ need to minimize the condition’s seriousness. These studies also reported that false positive subjects showed a confirmatory search strategy, in which they recalled symptoms and risk behaviors that fit the new health condition label better than their false negative counterparts. There were no differences in recall for general (nondiagnosis relevant) symptoms (Croyle & Sande, 1988). Baumann, Cameron, Zimmerman, and Leventhal (1989) reported a similar bias when individuals were given false feedback about their blood pres­ sure. Those who believed their blood pressure to be high reported more symp­ toms commonly associated with hypertension, especially when they attributed hypertension to stress (see also Bishop, Briede, Cavozos, Grotzinger, & McMa­ hon, 1987). This result is interpreted in terms of a selective retrieval of symp­ toms that is afforded by one’s schema of the illness or condition (see Fig. 5.1).

250

CLARK

These findings parallel schema-driven retrieval effects in other area of social cog­ nition (Hastie, 1981; O’Sullivan & Durso, 1984; Wyer, Srull, Gordon, & Hart­ wick, 1982). Framing Biases and Other Contextual Factors Influencing Interpretation

It has been known for some time that the interpretation of symptoms and health conditions are influenced by many external factors (Pennebaker, 1982; Skelton & Pennebaker, 1990). Figure 5.1 depicts surrounding context information as an important factor in determining how stimulus information is interpreted and encoded. A major contextual factor concerns how diagnosis information is framed. The way in which information is presented strongly directs one’s information processing and responses (Cioffi, 1993; Kahneman & Tversky, 1984). Framing effects have been reported for health-relevant behavioral intentions regarding condom use (Lin­ ville, Fischer, & Fischhoff, 1993) and breast self-exam (Meyerowitz & Chai­ ken, 1987). Specifically, college students reported higher intentions to use a condom when it was presented as having a 90% success rate (success frame) ver­ sus a 10% failure rate (failure frame) at protecting them against AIDS. Meyerowitz and Chaiken documented the beneficial effects of presenting the negative conse­ quences of not performing breast self-exam (a loss frame) over presenting the gains of performing the behavior (a gain frame). The authors argued that a loss frame is needed to motivate risk-seeking behaviors, such as facing the possibility of being diagnosed with cancer. Cioffi (1991b, in press) studied frame effects in the actual presentation of di­ agnosis information. She argued that most diagnostic test results are framed in terms of the presence of pathology (feature positive), rather than its absence (fea­ ture negative). Research from social cognition indicated that individuals are less accurate and have more difficulty when diagnosis information is presented in terms of something that is absent or has not occurred yet (Fazio, Sherman, & Herr, 1982; Lord, Lepper, & Preston, 1984). Cioffi argued that such “wellness” or feature-negative diagnoses will pose particular problems when such test results are described as “uncertain” (Fazio, 1987; Tversky, Sattath, & Slovic, 1988), because individuals are forced to extract information from an uncertain nonoc­ currence of disease indicators. In comparing uncertain diagnoses about an en­ zyme deficiency, the uncertain wellness group perceived the disease as more severe, perceived themselves to be more at risk, and as more likely to have the disorder after a second identical test result compared with the uncertain “disease” group (Cioffi, 1991b). In addition, the uncertain wellness diagnosis appeared to block self-serving biases, such as minimization, used by individuals who received certain disease diagnoses. Framing a diagnosis in terms of an uncertain featurenegative hypothesis may lead individuals to continue to feel worried and at risk

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

251

even when they have received a reassuring diagnosis of wellness (Cioffi, in press). Another contextual factor that plays a role in how individuals immediately ap­ praise their own health diagnosis is the response of a comparison other (see so­ cial comparison section). The role of stress in both the labeling of sensations as symptoms and the labeling of symptoms is another critical contextual factor (Cacioppo, Anderson, Tumquist, & Tassinary, 1989; Cohen & Williamson, 1991; Leventhal, Meyer, & Nerenz, 1980). Cohen and Williamson presented evidence for a model in which stress heightens the salience of a sensation and, through the activation of specific schematic knowledge structures, affect the interpreta­ tions of those sensations. Finally, Cioffi’s (1991a) cognitive-perceptual model of somatic interpretation highlights the importance of individuals’ knowledge structures, goals, health hypotheses, and illness beliefs in determining how they appraise both symptoms and health conditions. She has also argued that the surrounding context in which the sensations are experienced greatly influences both the amount of attention paid to them and individuals’ subsequent interpretations of the sensations.

MENTAL REPRESENTATIONS RELEVANT TO HEALTH AND ILLNESS

As the previous section and Fig. 5.1 indicate, the mental representation of one’s illness knowledge plays an important role in the appraisal of somatic states (Croyle & Barger, 1993; Skelton & Croyle, 1991). Such representations also can influence one’s help-seeking behaviors (Prohaska, Keller, Leventhal, & Leventhal, 1987) and one’s ability to understand health communications (Burbach & Peterson, 1986; Morgan, Fischhoff, Bostrom, Lave, & Atman, in press). This section discusses research on the health-relevant mental representations of symptoms, illnesses, the self, and health threats. Representations of Sensations and Symptoms

Work on symptom representation has suggested that bodily sensations are represented mentally in terms of their physiological elements and the meanings that people attach to those elements (Pennebaker, 1982). Leventhal (1982) sug­ gested that individuals can create a representation of only the objective features of a bodily sensation without also encoding its higher level implications by focusing on the sensation in an emotionally detached manner. Such encoding facilitates distress reduction primarily by the formation of a schema that downplays or omits the threatening emotional implications that are so characteristic of new and un­ pleasant sensations. This notion of an objective schema was reviewed in the last section.

252

CLARK

If representations of physical sensations in terms of symptoms exist, then in­ dividuals should be able to prime the activation of them. In a study of symptom suggestibility, Skelton (1987) used a somatic imagery task to examine the impact of triggering the activation of symptom schemas on individuals’ experiences of physical symptoms. Specifically, he compared somatic, visual, and auditory im­ agery tasks. Subjects were shown word cues and were told to form a correspond­ ing image (e.g., cold hands). After each image, subjects rated the image, their mood, and the degree to which they were experiencing each of several physical symptoms. Subjects in the somatic group tended to report experiencing all physi­ cal symptoms, not just those that they had imaged, to a greater extent than the other image groups. However, several of these symptoms were related to the somatic cue words. In a second, within-subjects study, somatic imaging increased reports of imaged and image-relevant symptoms, but not irrelevant symptoms, compared with visual and auditory images (Skelton, 1987). Furthermore, sub­ jects who were most likely to report forming vivid images also reported experienc­ ing the greatest number of image-relevant symptoms. Skelton (1987) suggested a “matching to standard” process to account for these findings, in which individuals compared their current bodily state to their represen­ tation of the symptom to rate the degree to which they were experiencing the symptom. Activation of this representation through imaging led to exaggerated perceptions of actually experiencing the immediately imaged symptom. In effect, Skelton argued, accessing the representation that was recently activated led to a projection of the sensation to the bodily locus associated with the symptom, thereby creating the bodily sensation that the individual was currently rating. Such a finding may shed light on the experience of hypochondriacs. That is, these per­ sons might actually experience those sensations that they constantly worry about as symptoms of illness. It could also explain why imaging of physical states used in coping produces beneficial relaxation effects. Representations of Illnesses

Most of the work on health-relevant representation has focused on individuals’ beliefs and experiences relevant to disease or illness. In evaluating sensations, symptoms, and illness signs, individuals rely on information that is accessed from long-term memory in the form of schemata (Cacioppo et al., 1989). These sche­ mata hold individuals’ knowledge and beliefs about the physiological outcomes associated with health and illness conditions. Once activated, these knowledge structures lead to selective attention, anchoring effects in judgment, and a con­ firmatory search strategy, just as in other domains of social cognition (Fiske & Taylor, 1991; Hastie, 1981). Leventhal (1986) suggested that illness informa­ tion is actually represented in three distinct forms; physical-motor sensation com­ ponents, perceptual schemas, and conceptual codes.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

253

Conceptual Knowledge About Diseases. Conceptual knowledge includes declarative knowledge descriptions (Rumelhart & Ortony, 1977) that character­ ize a particular disease or illness. Such representations of illness contain knowledge in the form of four components: identity of illness (label and associated symp­ toms), cause of illness, time line, and consequences. In a study of the use of such knowledge representations, Prohaska et al. (1987) asked subjects for their emo­ tional and coping responses to a series of scenarios that involved common sets of symptoms varying in severity and duration. They found that responses to scenarios involving more severe symptoms showed more emotional distress and a higher incidence of self-care and medical care seeking behaviors. Scenarios depicting symptoms of longer duration also elicited medical care seeking behaviors. When individuals attributed the cause of the symptoms to aging and not illness, they showed less emotional distress and tended to delay in seeking treatment. Bishop (1987) also found that perceptions of causes affected intentions for health behaviors. Specifically, symptoms attributed to a viral cause initiated self-care behaviors, whereas seeing the cause as more physical than psychological in na­ ture generally increased professional care seeking. Contextual cues have a great­ er impact on the interpretation of symptom sets when the individual has less familiarity with an illness, and hence a less well-developed representation of the corresponding illness (Baumann et al., 1989). Research by Lau, Bernard, and Hartman (1989) and Lau and Hartman (1983) detailed evidence for Leventhal’s four representational components and added a fifth component in the form of cure information (Bishop et al., 1987). Lau and Hartman (1983) asked subjects to describe their knowledge of everyday diseases and illnesses. Subjects spontaneously described their knowledge in terms of the label, a time line, consequences, causes, and cures. In addition, their answers about the reasons for getting sick (cause) and getting well (cure) were associated with changes in beliefs about their personal control. Specifically, controllable attributions for getting sick and internal attributions made for getting better were associated with increases in beliefs about self-control over their health. On the other hand, attributing the cause and recovery to unstable reasons increased their belief in chance health outcomes (see also Lau et al., 1989). Lau et al. (1989) showed the stability of the five components using a free cluster­ ing technique of subjects’ spontaneous descriptions of illnesses. Subjects who showed strong identity and cure components also reported a higher likelihood of visiting a doctor when feeling ill. In sum, converging evidence has suggested a general consensus for the five component taxonomy of representational com­ ponents for generic illness. However, using a factor analytic strategy, Turk, Rudy and Salovey (1986) did not find straightforward support for the five-component model. Nurses, dia­ betics, and college students responded to a variety of questions about diabetes, cancer, and the flu. The important dimensions of illness were seriousness, personal

254

CLARK

responsibility (for cause and cure), controllability, and changeability. In addi­ tion, personal experience with a disease affected subjects’ responses. For exam­ ple, diabetics and nurses specializing in diabetes showed similar weightings that differed from college students’ conceptions of diabetes. Overall, Turk et al. con­ cluded that these dimensions represented a general implicit model of illness. Other researchers have suggested that the divergence of these findings from the previous research was due to differences in research approaches. Specifical­ ly, the findings of Turk et al. (1986) focused on the differences among different disease conceptualizations, whereas the component approach focused on shared features of disease (Bishop, 1991; Croyle & Barger, 1993; Lau et al., 1989). Turk et al. suggested that their findings show the subjective appraisal of individuals’ experiences with a disease, including the personal and affective consequence. In fact, Leventhal (1986) defined individuals’ perceptual schemas for illness as some­ thing different than their abstract conceptual knowledge. This distinction is im­ portant for studying how and when individuals use their personal experience with an illness, rather than their general knowledge, to determine their actual responses to symptoms. Episodic Exemplars Versus General Prototypes. Perceptual schemas hold in­ formation regarding symptoms, emotional reactions, contextual events, and specif­ ic illness episodes. These elements are associated through the co-occurrence of these components in individuals’ experiences. Such schemas share many features with exemplar-based representations, which have been talked about in cognitive psychology (Hintzman, 1986; Smith & Medin, 1981) and social cognition (Smith & Zarate, 1992). Illness episodes are activated through a template-matching process of currently experienced symptom sets to the schema. If the person’s cur­ rent bodily sensations (symptoms) match past experienced illness episode, then he or she will use the illness episode as his or her self-diagnosis (Leventhal & Diefenbach, 1991). In contrast to the idea of perceptual schemas or exemplars, Bishop (1991) and Bishop and Converse (1986) suggested that illness beliefs are represented in terms of more abstract prototypes. These prototypes are conceptualizations that influence the identification of disease, the recall of specific illness experiences, and the processing of information about those experiences. For identification purposes, a prototype serves as a standard against which individuals match up information about symptoms. Activation of a particular prototype does not depend on an ab­ solute match, but rather a close resemblance. However, the closer the fit of stimuli to the prototype, the greater facilitation it affords to later recall of the symptom information. Bishop and Converse (1986) gave subjects descriptions of 12 stimulus in­ dividuals, along with information that these individuals were experiencing vari­ ous combinations of symptoms. These symptom sets varied in their prototypical­ ity to different diseases. Subjects were more likely to say that a target individual’s

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

25 5

symptoms indicated a disease when the target exhibited a symptom set that was highly prototypical of a disease, versus when the symptoms were only moderate or low with respect to prototypicality. In addition, subjects made a correct or related diagnosis more often when the symptoms were high in prototypicality com­ pared with the other types of symptom sets. In a second study, Bishop and Con­ verse showed that accurate recall was greater for symptom sets that were high in prototypicality compared with sets low in prototypicality or a random combi­ nation of symptoms. Finally, in another study, subjects were faster in judging the representativeness of symptoms that were high in prototypicality to a specific disease compared with those that were not as prototypical (Bishop, 1991). As in the case of person perception more generally, both exemplar and proto­ type representations of illnesses are likely to exist. It should be noted that Bishop’s studies used a task in which subjects read about different individuals who ex­ hibited specific symptom sets. There may be some differences in the way in which individuals apply labels to the experience of others versus themselves. The next section examines work on individuals’ conceptualizations of the types of people who are likely to experience particular illnesses. Representations of Individuals Experiencing an Illness

Illness stereotypes have been defined as beliefs about people who have a particu­ lar disease (Croyle & Williams, 1991). When individuals are told that they have an illness that is associated with positive personality traits, they may feel less threatened and may be less likely to appraise the condition as being a serious health threat, than when the illness stereotype has negative personality connota­ tions (Croyle & Williams, 1991). In addition, such stereotypes may affect in­ dividuals’ reports of health behavior and symptoms associated with the illness. In this regard, Croyle and Williams found that high blood pressure was believed to be associated with positive personal characteristics (e.g., “good grades”) than with negative characteristics (e.g., “gives up easily”). Moreover, the more posi­ tive the stereotype that individuals held, the less serious they rated the condition. In a second study, Croyle and Williams (1991) manipulated subjects’ beliefs about the type of people who experience hypertension. They also gave false feed­ back to subjects regarding their own blood pressure. The positivity of the stereo­ type correlated negatively with subjects’ appraisal of the seriousness of hypertension. In addition, subjects who believed that they had exhibited high blood pressure, but were given positive stereotype information, reported fewer blood pressure-irrelevant symptoms than other subjects. Correspondingly, subjects who believed they had high blood pressure and received negative stereotype informa­ tion reported fewer blood pressure-relevant symptoms than other subjects. These results suggest that subjects’ information search processes were biased in an at­ tempt to disconfirm a self-disparaging diagnosis. Croyle and Williams (1991) con­ cluded that it is important to incorporate perceivers’ illness stereotypes into models

256

CLARK

of health-threatening appraisals. In addition, they suggested that an understand­ ing of the perceptions of disease victims will be incomplete without a knowledge of perceivers’ beliefs about the disease. AIDS is a medical condition that, due to its serious consequences regarding recovery, may lead individuals to shy away from individuals who are known to have the disease. This is an instance where perceivers’ knowledge of the disease may affect their reactions to others with the disease. Evidence of the impact of individuals’ knowledge structures on their emotional reactions was shown by Heath, Acklin, and Wiley (1991). They asked physicians who treated AIDS pa­ tients to imagine contracting AIDS from treating a patient. Doctors who mental­ ly simulated this outcome later showed increased worry and concern regarding this particular risk. Fish and Rye (1991) found that subjects’ knowledge of AIDS was positively correlated with their evaluations of someone with AIDS, but was not related to their desire to interact with such a person. In an attempt to explain discriminatory responses to AIDS victims, Pryor and Reeder (1993) examined individuals’ representations of individuals who experience AIDS. In Pryor and Reeder’s representational framework, AIDS was considered to be an attribute that is linked with a given person node in an associative net­ work structure, much like those proposed by researchers in cognitive psycholo­ gy (Anderson & Bower, 1979; Collins & Loftus, 1975) and social cognition (Carlston & Skowronski, 1986; Wyer & Carlston, 1979). These structures con­ tain perceivers’ beliefs stored in interconnected propositional units, as well as associated affective responses that can be triggered when the representation is activated (Fiske, 1982). The strength of the links between these units or nodes is a function of individuals’ cultural and idiosyncratic experiences. Pryor and Reeder argued that affect may reside in several sources. For exam­ ple, a person’s representations may have affect if it is stored with various infor­ mational nodes associated with AIDS, such as its life-threatening consequences, drug use, or homosexuality. Pryor and Reeder also instilled procedural proper­ ties in their representation, so that a chain of nodes that has been repeatedly ac­ cessed can become so associated as to be triggered as a unit. For example, nodes regarding homosexuality (and its associated affect) may be triggered automati­ cally when the person encounters someone with HIV. The affective response that a given perceiver experiences is a sum of all the affect sources that have been activated through spreading activation processes in the network. Evidence for an AIDS representation comes from work on social judgment studies that involve AIDS targets (Pryor & Reeder, 1993). Specifically, negative attitudes toward homosexuality adversely affected business employees’ reports of willingness to hire someone with AIDS, even though the target was never iden­ tified as being homosexual. Similarly, individuals with negative attitudes toward homosexuality reported more negative reactions when asked to respond to a scenario in which their children’s classmate has AIDS (Pryor, Reeder, & McMa­ nus, 1991).

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

257

Pryor and Reeder replicated these findings using descriptions of a fictitious disease. Subjects were told that one contracted this disease primarily from eating shellfish. Half of the subjects also was told that this disease could, at times, be sexually transmitted and had infected some homosexual men. Subjects who held antihomosexual views and had received this additional information about the dis­ ease were most likely to reject a target person with the disease, even though the target was not identified as being homosexual. Finally, Pryor and Reeder (1993) used subliminal priming (Bargh & Pietromonaco, 1982), in which the concept of homosexuality was activated be­ low the level of consciousness in half of the subjects. Afterward, subjects rated their reactions to having a roommate who suffered from a noncontagious, un­ usual disease. When homosexuality had been primed, subjects’ reactions were correlated with their attitudes toward homosexuality. Negative attitudes toward homosexuality led to more rejection of the target, even though he was not identi­ fied as being homosexual. Affect associated with homosexuality apparently was activated during the priming task and, in effect, spilled over onto perceivers’ per­ ceptions of the potential roommate. Pryor and Reeder suggested that these studies demonstrated that nonconscious representational links of AIDS to homosexuality is, in large part, responsible for the stigmatization that AIDS victims experience (although see Bishop, Alva, Cantu, & Rittman, 1991). Representations of the Self and Implications for Illness

The next section on representation examines the role that self-representations play in affecting one’s likelihood of becoming ill or in adapting to health disorders. Self-Represented Knowledge of and Vulnerabilities to Illness. Linville (1987) argued that individuals store self-knowledge in terms of domain-specific self­ aspects. These aspects may share among them many connections, representing the interdependence individuals view between their various self-aspects. Individu­ als with high self-complexity used many self-aspects that were relatively inde­ pendent of one another. Individuals with low self-complexity used only a few overlapping constructs or self-aspects. Linville found a relationship between the degree of complexity of individuals’ represented self-knowledge and the likeli­ hood that stress would lead them to become ill. Individuals with high self-complexity were less likely to experience stress, phys­ ical symptoms, flus, and other illnesses in response to stressful life events than were individuals with low self-complexity. Presumably representing self-relevant information in terms of a high number of independent of self-aspects served to buffer these high self-complex individuals from stress. Linville explained this finding in terms of individuals’ reactions to stress. Those high in complexity enjoyed relatively compartmentalized self-concepts and at­ tributes. Therefore, when a negative life event affects an individual’s component

258

CLARK

of self-knowledge, other components are still able to provide evidence of an in­ dividual’s worth and the positivity of his or her life. When one identifies him or herself in terms of few, intertwined aspects, a negative life event can lead to ruminations and negative implications across all self-aspects, increasing vulner­ ability to the negative effects of stress. Representations of individuals’ self-relevant standards can also leave them vul­ nerable to health-aversive behaviors (Higgins, Vookles, & Tykocinski, 1992; Strauman, Vookles, Berenstein, Chaiken, & Higgins, 1991). Specifically, research on self-discrepancies has suggested that individuals experience discomfort due to a mismatch in perceptions of this actual self and some standards concerning what they would like to be (ideal self) or what they should be (ought self). These standards may consist of their own criteria or the perceived criteria of significant others. Self-relevant perceptions are represented in long-term memory both in­ dividually and in terms of discrepancies (e.g., actual:ideal discrepancy). When activated through usual representation activation processes, such as priming, these discrepancies influence individuals’ emotions and behaviors (Higgins, 1987). Strauman et al. (1991) found that certain discrepancies differentially predict­ ed attitudes and behaviors relevant to eating disorders (see also Forston & Stan­ ton, 1992; Higgins et al., 1992). Based on DSM III descriptions (American Psychiatric Association, 1980), they hypothesized a relationship between bulim­ ic behaviors and failure or helplessness, and a relationship between anorexic be­ haviors and feelings of guilt. They identified individuals who held self-concepts entailing feelings of unfulfilled positive potential, represented as a form of actu­ al: ideal discrepancies that focused on self-possibilities (the “can” self). This type of individual was predicted to be more likely to engage in bulimia-related be­ haviors due to the experience of feelings of failure and a lack of control. They also identified individuals who did not perceive themselves as living up to the standards of how others thought they should be. Individuals experiencing these actual own:ought other discrepancies were predicted to engage in anorexic be­ haviors due to feelings of guilt and the need to live up to the expectations of others. Strauman et al. (1991) found that individuals with the “can” self form of actu­ al: ideal discrepancy showed greater self-reported preoccupation with food and binge eating behaviors, whereas those with the actual own:ought other discrepancy reported higher dieting and oral control attitudes and behavior (see also Higgins et al., 1992). Representations of the self in terms of individuals’ own self-expectations or desires and the perceived expectations of others have important influences on af­ fect, and health-relevant attitudes and behaviors. These studies indicate that in­ dividuals respond to specific self-discrepancies with particular behaviors that serve to threaten their physical well-being. In addition, Strauman and his colleagues have demonstrated that priming self-standards can reduce the natural killer cell activity in clinically anxious and in dysphoric individuals, thereby negatively af­ fecting their immune responses (Strauman, Lemieux, & Coe, 1993).

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

25 9

This research provides examples of the function of self-representations in decreasing or increasing people’s vulnerability to health problems. At times, a serious medical condition can impact self-concept negatively. The following is an example of how self-representations may undergo changes as part of the process of adjustment to a long-term medical condition. Changes in Self-Representation as Adaptation to a Long-Term Medical Con­ dition. Clark, Henry, and Taylor (1991) concluded from the infertility litera­ ture that the experience of infertility represented the blockage of a major life goal. Furthermore, the major life goal of having a child was classified as an inherited goal (i.e., a goal that is passed down culturally through socialization). For many, such a goal remains relatively unquestioned. Due to the inherited nature of the goal to have a child, it may be represented vaguely as embedded (seen as syn­ onymous) with adulthood fulfillment. Much of the psychological experience of infertility revolves around adjusting to the consequences produced by goal blockage or frustration. Information represented in the framework with parenting goals eventually may be questioned by infertile individuals. This questioning process is initiated due to the (a) threat of failing to achieve parenthood, (b) emotional and physical costs of continued treatment, and (c) social stigma imposed by others (Clark et al., 1991). Questioning one’s goals can lead to an elaborated representation by mak­ ing clear distinctions between potential fulfillment of certain aspects of parenting and by disentangling the goal from other high-level aspirations (i.e., to be a ful­ filled individual). The cognitive reconstruction that takes place as a function of questioning why one wants to be a parent makes an individual more aware of which subgoals are potentially achievable and which ones are not. Unachievable aspects of parent­ ing (e.g., having one’s own biological child) are mourned. Overall, however, positive adaptive benefits accrue from this process, because the individual feels an increase in awareness of subgoal distinctions, increased personal control, and becomes aware of an increased number of options for happiness. This restructur­ ing process also allows an individual to reinvest effort in other aspects of life (e.g., work, marriage). Thinking in terms of a restructuring of embedded goals provides a framework for suggestions to therapists and social support members on how to conceptualize the cognitive experience of the distressed infertile per­ son. The next section reviews research involving individuals’ understanding of information regarding potential health threats. Capturing Individuals' Mental Models of Health Risks

A major goal of communicators of health risk information is to provide informa­ tion in a way that fits individuals’ intuitive ways of thinking. Without taking into account individuals’ representations, information that is disseminated by health

26 0

CLARK

officials may be ignored or misinterpreted. This work again points to the role of preexisting knowledge impacting on the interpretation of incoming stimuli as depicted in Fig. 5.1. Morgan et al. (in press) developed a technique to uncover the layperson’s con­ sensual knowledge of environmental health threats. This knowledge is represent­ ed in terms of a mental model of individuals’ beliefs. Beliefs are first elicited through open-ended interview questions. Next, these beliefs are categorized in terms of content areas relevant to (a) exposure, (b) effects, (c) mitigation be­ haviors, (d) evaluation, and (e) background information. In addition, beliefs are coded as being accurate, inaccurate, peripheral, or indiscriminate. Morgan et al. then described how this technique is used to elicit the represen­ tations of health risk experts. In designing communications, Morgan et al. used the experts’ representation combined with the target audiences’ erroneous, peripher­ al, and background beliefs. In this way, they tailored the communication in ways that would enable them to refine the mental models of the audience. Brochures about radon exposure designed according to this procedure were compared against brochures designed by the Environmental Protection Agency (EPA). These two brochures yielded no differences in memory for the material. However, subjects who read brochures based on empirically derived mental models showed better performance on questions that required them to make inferences on topics not explicitly mentioned in the text, compared with the EPA brochure group. In addition, these subjects gave more detailed recommendations about how to handle the health threat. In sum, the approach that took into account individu­ als’ mental models in preparing communication appeared to be efficacious in in­ creasing individuals’ understanding of a health threat. This section focused on the knowledge that individuals have stored in memory about illnesses, illness stereotypes, and health threats. Self-relevant constructs in the form of the self, standards, goals, and complexity of self-knowledge all impact on individuals’ current health and ability to adapt to health problems. The next section explores the influence of affect in the form of mood and personality predispositions in people’s processing of health-relevant information.

THE ROLE OF AFFECT IN HEALTH COGNITION, MEMORY, AND ADJUSTMENT

Clearly, affect is influenced by health and illness concerns. The role of affect as an outcome in adaptation to chronic illness conditions is beyond the scope of the present chapter. However, as Fig. 5.1 indicates, affect influences informa­ tion processing in the form of antecedent person and situational factors. Specifi­ cally, mood and affective orientations can determine the way in which individuals interpret stimuli, retrieve relevant information, and respond to health-relevant concerns.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

261

The Role of Affect in Health-Relevant Judgments and Memory

Research in health psychology has studied memory bias and distortions across a wide variety of phenomena, including the retrospective reports of pain (Salovey, Sieber, Jobe, & Willis, in press), AIDS-related risk behaviors (Croyle & Loftus, in press), and reasoning behind failed treatment decisions (Wagener & Taylor, 1986). However, the issues of this section focus specifically on the in­ fluence of mood and affect on individuals’ memory for health-relevant experiences. Mood and Recall o f Pain, Physical Symptoms, and Health-Relevant Be­ haviors. In a study of pain perception during labor, Norvell, Gaston-Johansson, and Fridh (1987) found that, 2 days postpartum, women rated labor pain as be­ ing significantly less intense than they originally had rated it during the three stages of labor. This finding was interpreted in terms of a mismatch between the mood state of women during labor (e.g., anxiety) and the moods they experienced 2 days later (e.g., joy, relief). Similar findings were reported for rheumatoid pa­ tients’ pain ratings prior to and after surgery (Roche & Gijsbers, 1986). Patients rated their presurgery pain as less intense once surgery was over than they had rated it initially. Again, the conclusion is that relief after surgery distorted presur­ gery pain recall. Finally, individuals high in dental fear recalled their dental pain as being more intense than they actually rated it at the time they experienced it (Kent, 1985). Memory techniques have been used that employ a focus on episodic details to improve health-relevant recall (Croyle & Loftus, 1992; Salovey et al., in press). For example, in a study of dietary recall, Fisher and Quigley (in press) asked half of their group of female subjects to remember the emotional and physical details of a party they had attended 2 weeks previously. This recall of past emo­ tions associated with the event increased recall accuracy of the foods they had eaten at the party. Eich, Rachman, and Lopatka (1990) conducted a study examining the impact of current pain on the accessibility of episodic memories. They found that effects of pain to facilitate recall of unpleasant experiences and inhibit recall of pleasant experiences was mediated entirely by affect. That is, only when pain was accom­ panied by negative affect did it influence the retrieval of past experiences. The previously cited studies were suggestive of a mood and memory link, but manipulations of mood provided stronger tests. Croyle and Uretsky (1987) manipu­ lated subjects’ mood by showing them a sad or happy film. Afterward, in an un­ related task, subjects rated the physical symptoms they had experienced during the last month. Seeing a sad film led to higher reports of physical symptoms, whereas seeing a happy film led to lower reports than a no-film control group. Salovey and Birnbaum (1989) replicated this finding using an autobiographical mood induction technique with college students who were experiencing the flu. Mood effects were particularly pronounced on flu symptoms that indicated pain,

262

CLARK

compared with nonpain flu symptoms. Finally, Salovey et al. (1992) induced hap­ py, sad, or neutral moods and then asked subjects to recall a painful incident from the past. They rated this episode for pain intensity and also rated stories of hypothetical pain episodes. Subjects’ ratings were influenced by their mood; namely, a sad mood tended to increase ratings and a happy mood decreased pain somewhat relative to neutral moods. Four Mechanisms for the Mood and Memory Link. Four explanations of the previous mood effects on memory were suggested by Salovey et al. (in press). First, a mood-congruent reporting explanation suggests that an individual’s mood at the time of retrieval will differentially affect the accessibility of memories that are similar in valence or affect. Second, a difference in the affect experienced at encoding versus retrieval will lead to biases in recall, such as the decrease in pain intensity recalled by new mothers and postsurgery rheumatoid sufferers (Norvell et al., 1987; Roche & Gijsbers, 1986). State-dependent recall facilita­ tion is blocked by this mismatch in affective context (Bower, 1981; Bower & Meyer, 1989). A third explanation suggests that increased reports and ratings of pain due to depressed moods may be attributable to an increase in self-focus brought about by the negative mood (Pyszczynski & Greenberg, 1987; Salovey, O’Leary, Stretton, Fiskin, & Drake, 1991). Fourth, a mood-as-information model suggests that one’s mood at the time of recall (or judgment) may bias his or her perceptions, because individuals use their present mood as input into their judgments about past experiences (Clore, Schwarz, & Conway, chapter 7, volume 1; Schwarz, 1990). Specifically, individuals who are unaware of the source of their present mood may misattribute their current positive or negative feelings to the stimulus they are being asked to rate (e.g., the pain or physical symptoms experienced of last week). Therefore, current happy moods lead to decreased ratings of past pain intensity and current sad moods elevate one’s ratings of past pain intensity (Salovey et al., in press). Mood and Other Health-Relevant Perceptions. Health-relevant perceptions other than symptoms also are affected by moods. For example, individuals’ per­ ceptions of efficacy were influenced by mood (Salovey & Birnbaum, 1989). Spe­ cifically, ill subjects in a sad mood rated themselves as less able to carry out illness alleviating behaviors and were less confident that these behaviors would be ef­ fective than happy subjects. Clark, Garstka, Bless, and Shah (1992) examined the effects of manipulated mood on individuals’ self-perceived coping styles. Happiness lowered scores on a confronting coping style scale, whereas negative mood increased confronting style scores. This finding was interpreted as a mood as information effect on self­ perception (Schwarz, 1990). In fact, this effect of mood disappeared when sub­ jects were made aware of the induction task’s influence on their moods.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

26 3

Finally, the affect of a physician presenting mammogram test results influenced individuals’ recall and judgments (Shapiro, Boggs, Melamed, & Graham-Pole, 1992). Videotapes were shown of an oncologist presenting ambiguous mammo­ gram results with either worried or nonworried affect. Subjects who were at risk for breast cancer were asked to imagine that they had taken a mammogram and were hearing the results of their own tests. Subjects in the worried physician con­ dition recognized significantly less information from the tape and rated the clini­ cal situation as more severe than subjects in the nonworried physician condition. These subjects also showed higher anxiety after viewing the tape, both in terms of pulse rate and self-reports. Mood clearly affects individuals’ judgments and recall of health-relevant in­ formation. Mood normally is thought of as a transitory affective state. Affect, as a dispositional variable, also influences health-relevant cognition and behavior. The next section reviews work on the relationships between affective orienta­ tions, such as optimism, hostility, and health. Stable Affect-Relevant Orientations Toward Information Processing

Work on chronic dispositions that individuals exhibit in responding to stress and health threats has suggested that such personality styles impact on both percep­ tions of and responses to health threats. For example, cancer patients dispositionally high in anxiety are more likely to develop anticipatory nausea in response to chemotherapy treatment (Van Komen & Redd, 1985). This section examines research for three individual differences regarding affective orientations on in­ formation processing that appear to heighten or lower individuals’ likelihood of becoming ill. These are optimism, negative affectivity, and hostility. Optimism. Researchers have identified ways in which positive emotions may serve adaptive functions (Lazarus, Kanner, & Folkman, 1980; Levy, Lee, Bagley, & Lippman, 1988). First, positive emotions allow individuals a break from the arduous demands of coping, thus acting as a “breather.” Second, positive emo­ tions in the form of hope and anticipation can act to “sustain action” in the face of taxing and demanding situations. Third, positive emotions can serve to “re­ store” depleted resources by providing relaxation and so on. The second of these, emotion as a sustainer of action, relates to the personality disposition of opti­ mism (Scheier & Carver, 1985, 1992). Specifically, Scheier and Carver argued that an individual’s actions are influenced by his or her expectancies regarding the consequences of those actions. Optimists generally expect success or positive outcomes, and therefore react to threatening situations in a planful (and positive) manner. Optimists report being less bothered by physical symptoms (Scheier & Car­ ver, 1985) and less emotionally disturbed by breast cancer surgery (Scheier &

264

CLARK

Carver, 1992) or testing positive for HIV (Taylor et al., 1992). In a study of coronary by-pass patients, Scheier et al. (1989) found that optimistic individuals showed less depression prior to the surgery and more relief and happiness after­ ward than their pessimistic peers. They also showed more evidence of planful behavior by seeking out information concerning what was to be expected of them in the months ahead and setting goals for recovery. Consequently, these individuals showed more satisfactory recovery as indexed by both staff and self-ratings. Fi­ nally, optimists showed greater normalization in their life activities 6 months af­ ter the surgery, compared with pessimists. These results are not unlike those reported in studies of individuals who chose to attend to a stressful situation, rather than avoid it (see section on attentional processes). There is also evidence that, besides engaging in a planful outlook, optimists behave in ways that are health promoting. Optimistic undergraduates were more likely to engage in health maintenance behaviors than pessimists (Robbins, Spence, & Clark, 1990). Optimistic coronary rehab patients exhibited behaviors that produced reductions in coronary risks (e.g., exercise, diet), compared with pes­ simistic counterparts (Marato, Sheppard, & Pbert, 1990; Scheier & Carver, 1992). Finally, optimistic individuals who tested negative for HIV reported having few­ er anonymous sexual partners than pessimists who tested negative (Taylor et al., 1992). In sum, optimists appear to fare better than pessimists both in the experience of discomfort and distress and in long-range benefits construed through action and planning. The concept of optimism clearly is related to personality constructs of hardiness (Allred & Smith, 1989; Matthews, 1988), a sense of personal con­ trol (Rodin & Salovey, 1989; Taylor, 1991), and enthusiasm (Friedman, 1993) — three other constructs related to the way in which people perceive, interpret, and respond to potentially stressful stimuli. However, the story is not so simple. Although optimists actually seem to be better than pessimists at accepting unchangeable aspects of negative health situa­ tions (Scheier & Carver, 1992), others have argued that too much optimism might not always be a good thing. Researchers have questioned whether too much opti­ mism may be unrealistic and lead individuals to avoid dealing with potentially dangerous health signs (Weinstein, 1982). Before being confronted with actual explicit threats, individuals in general tend to believe that they are relatively im­ mune or invulnerable to health threats that others will experience. For example, college students perceive their risk of acquiring an HIV infection through sexual contact as significantly lower than the average college student (Linville, Fischer, & Fischhoff, 1993). This finding has been called the optimistic bias (Weinstein, 1982). Two studies examined conditions under which such a bias may be reduced. First of all, individuals who are ill are less likely to show the optimistic bias for other health problems, but continue to show it for non-health problems (Kulick & Mahler, 1987a). This appears to be due to shift of attentional focus, induced by the current state of illness, to the lack of ability one has to prevent illness

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

26 5

in general. Individuals’ inability to prevent their current illness is generalized to other, unrelated health problems, eliminating individuals’ perceptions that they are less likely than peers to get sick. The optimistic bias also was eliminated when individuals were given responses of typical peers for each of the health risks. This study showed that the bias stemmed from a lack of attention regarding the probabilities of others, brought on by an egocentric approach to the task, rather than a lack of careful attention regarding their own vulnerabilities (Weinstein, 1983). One criticism of the optimism as a coping strategy work is that optimism scales merely capture the flip side of another personality construct referred to as nega­ tive affectivity (Smith, Pope, Rhodewalt, & Poulton, 1989). Although research has indicated that optimism effects cannot be accounted for totally by negative affectivity (Carver, Scheier, & Pozo, 1992), negative affectivity is an important construct. Negative Affectivity. Negative affectivity is defined as a stable personality dimension that consists of chronic negative emotions, such as anger, anxiety, guilt, insecurity, and sadness; and associated cognitive and behavioral characteristics, such as low self-esteem, preoccupation with worries, and a tendency to complain about things (Smith et al., 1989). Smith et al. found that negative affectivity ac­ counted for the same variance as optimism when studying the relationships among optimism, reports of physical symptoms, and a tendency toward problem solv­ ing, rather than passive coping. Although some research has found that negative affectivity affects only the reporting of health complaints, and not actual physical illness (Watson & Penne­ baker, 1989), links between negative affectivity and disease also have been de­ tected. In particular, Friedman and Booth-Kewley (1987) found weak, but consistent, associations between negative affectivity and asthma, arthritis, ulcers, headaches, and coronary heart disease. In addition, 25-year-olds who exhibited a pessimistic explanatory style, explaining negative life events in terms of inter­ nal, stable, and global factors, were in poorer health 20-35 years later, even af­ ter controlling for their initial health status, compared with individuals who did not exhibit this style (see also Antoni & Goodkin, 1988; Peterson, Seligman, & Vaillant, 1988). Negative affectivity can confound the relationship reported between stressful life events and illness, simply because individuals with this style receive high scores on both measures. Once contaminated items are removed from stress meas­ ures, the relationship between experienced stress and well-being drops dramati­ cally (Brett, Brief, Burke, George, & Webster, 1990). Brett et al. reviewed four possible mechanisms for this confounding effect. First, many items on life event scales are judgmental in nature (e.g., troubles with one’s spouse or boss). In­ dividuals high in negative affectivity may be more likely to interpret ambiguous stimuli as negative. Second, depressed mood can produce increases in reports

266

CLARK

of negative life events. Third, the experience of negative affect might provide these individuals with an encoding and retrieval advantage for unpleasant ex­ periences. Finally, negative affectivity may provide individuals with a negative life view that is responded to in ways that create adverse circumstances. Only the last explanation suggests that negatively affective individuals actual­ ly experience increased stress, rather than simply producing a response bias on these scales. Recent research on the specific predisposition of cynical hostility also has suggested that such individuals’ social responses play a role in increased stress and vulnerability to illness. Cynical Hostility. Research on Type A as an individual difference has found that one subcomponent, the potential for hostility, is linked to cardiovascular dis­ ease (see Dembroski, MacDougall, Costa, & Grandits, 1989; Matthews, 1988 for reviews). This component of “cynical hostility” is related to individuals’ health risks for cardiac problems (Smith & Pope, 1990) and even HIV infection (Blaney et al., 1991). Williams, Barefoot, and Shekelle (1985) found a relationship be­ tween increased physiological reactivity and a distrustful outlook toward others. In an effort to explain this relationship, they argued that hostile individuals are relatively more prone to experience anger in situations and to maintain a state of vigilant observation of their social environment. Hostile individuals also are more likely to experience interpersonal conflict, and to respond socially in ways that create stress and undermine social support (Smith & Christianson, 1992). Supporting the idea that hostile individuals react negatively in social situations, individuals scoring high on hostility reacted with higher blood pressure than those low in hostility when engaged in debating tasks (Smith & Allred, 1989), roleplaying of interpersonal conflict (Hardy & Smith, 1988), and situations involv­ ing self-disclosure about personal stress to a stranger (Christianson & Smith, 1991). In addition, Allred and Smith (1991) tested the idea that individuals high in cyni­ cal hostility hold a chronically accessible schema (Higgins & King, 1981) of “others as hostile.” This schema guides individuals’ perceptions and interpretations toward seeing others as threats and whose actions are driven by hostile intent. Allred and Smith (1991) hypothesized that hostile individuals would show selec­ tive attention and interpretation effects similar to past schematic effects in person perception (Hastie, Park, & Weber, 1984; Wyer & Gordon, 1984). They tested this hypothesis in three ways. First, hostile and nonhostile subjects rated an in­ teraction partner who was either (a) a hostile confederate or (b) a confederate whose actions were ambiguous with respect to hostility. Second, these subjects also read a paragraph about a target whose actions were ambiguous with respect to hostility. Finally, subjects were asked to recall the adjectives that were used in the two rating tasks. Hostile subjects rated their interactional partner as more hostile than did nonhostile subjects, irrespective of the confederate’s ambiguity. They also rated the hypothetical person in the ambiguous paragraph as more hostile than did nonhostile subjects. Recall findings showed an interaction between the

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

267

hostility of the subject and the hostility of the confederate. Specifically, hostile subjects recalled more hostile adjectives after interacting with a hostile confeder­ ate than any other condition. These subjects showed superior recall only for hostile adjectives, and not for negative adjectives in general. Finally, hostile subjects themselves acted more hostile in their interactions and rated themselves as more angry after their interactions than did nonhostile subjects, regardless of confederate type. Other work on the relationships among anger scales has suggested that the experience and expression of anger may be more important to the concept of poten­ tial for hostility than is the suspicious nature of such individuals (Musante, MacDougall, Dembroski, & Costa, 1989). In addition, hostile individuals may show increased health risks through their actual health-relevant behaviors. Specifical­ ly, hostile individuals report less physical activity, less self-care, more smoking, and more drinking than their low hostile peers (Koskenvuo et al., 1988; Leiker & Hailey, 1988). However, the work of Smith and his colleagues clearly pointed to the impor­ tance of cynical hostility as an individual difference that leads to potentially maladaptive social interactions and increased physiological arousal/distress in reac­ tion to social interactions. These reactions appeared to be mediated by the cogni­ tive style that such individuals use in interpreting others in a social context, as well as their own discomfort in social situations. In the next section, the health implications of others serving as sources for social comparison information are discussed.

SOCIAL COMPARISON AND HEALTH PSYCHOLOGY

Social factors play an important role in individuals’ appraisal of health status and adjustment to illness. Nowhere has this been made more apparent than in research on social comparison processes and health. In addition, according to Leventhal, Nerenz, and Steele (1984), social comparison provides the knowledge on which much of illness cognition, particularly symptom appraisal, is based. In terms of the information processing model used in this chapter, social com­ parison information may be thought of as (a) stimulus information that is sought or encountered; (b) surrounding context, such as the information about an ill­ ness’ prevalence; or (c) a coping response to a serious illness. Each of these ways of viewing social comparison becomes clear in the subsequent sections. Festinger (1954) first argued that individuals seek out information about the abilities and feelings of others to determine the relative standing of their own abilities and responses. People use others for social comparison when they lack objective criteria, and particularly in situations where they face uncertainty about what constitutes an appropriate response. In general, Festinger argued that in­ dividuals seek information about similar others who are doing slightly better.

268

CLARK

Making these upward comparisons was considered most informative for the pur­ poses of self-evaluation and self-improvement. In health psychology, the process of social comparison has been examined in the areas of symptom appraisal, reac­ tions to diagnosis, and overall adjustment to illness. Symptom Appraisal

In a chapter on social comparison and the seeking of nonexpert opinions, Sanders (1982) outlined why and how people seek advice, as well as how the adviser gener­ ates his or her opinions. Individuals seek the opinion of others regarding their health status, in part, because many somatic experiences relevant to the medical domain are complex and ambiguous, leaving them open to social definition. The cost of using medical experts is high in terms of money, time, effort, and embar­ rassment potential. Individuals may want opinions from friends and family mem­ bers as a way to deflect personal responsibility for the important decision of seeking treatment. Finally, physical impairment, combined with high levels of uncertainty, may produce a state of helplessness, in which individuals may be more likely to seek and depend on the guidance of others (Sanders, 1982). There are two ways in which individuals may seek to compare themselves with others. First, they may observe the behavior and experience of others. For ex­ ample, they may compare the frequency and intensity of salient others’ symptoms, or they may observe the others engaging in various health habits. The second way in which individuals gain social comparison information is through discus­ sion of their experiences with the salient others and gaining knowledge of those persons’ experiences. Areas in which comparisons of either type are likely to oc­ cur include symptom experience, health habits, and treatment. In the last case, information about others may help individuals establish criteria for monitoring their symptoms to determine when to seek professional care. In addition, infor­ mation regarding home remedies may be learned in this manner (Sanders, 1982). Sanders (1982) suggested that others provide information and opinions based on generalizing their own experiences. In a sense, advisers match up individuals’ complaints with an exemplar based on their experiences. If this match is faulty, bad advice results. In addition, part of the advisers’ general knowledge regarding symptoms leads them to advise the seeking of professional care for symptoms that interfere with daily functioning. Finally, Sanders noted that others are less personally threatened by the symptoms’ implications, and therefore are not moti­ vated to minimize the seriousness of individuals’ conditions. Because of this, they may be in a better position to regard potentially ill persons’ situations objectively. Reactions to Diagnosis Information

Although others may be objective, the work described earlier suggested that in­ dividuals are motivated to minimize a health threat (Croyle, 1990). However, the views of others regarding a health condition may affect individuals’ ability

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

269

to minimize the seriousness of a condition for which they have been diagnosed (Croyle, 1992). Croyle and Hunt (1991) manipulated both the diagnostic status and the response of a confederate who received feedback on an enzyme deficiency condition at the same time as the subject. Confederates either minimized the seriousness of the condition or made no comment. Subjects who tested positive showed less con­ cern and more avoidance of a second definitive follow-up test when the confederate made a minimizing comment. However, the long-term plans for dealing with the health threat were not affected by the confederate’s comments, but were influenced only by his or her diagnostic status. Subjects indicated less preventive health behavior intentions when the con­ federate tested positive, as well as the subject, than when the confederate tested negative. Presumably subjects in the former conditions believed the condition to be more common, and therefore less serious or threatening. Hence, the infor­ mation regarding the prevalence of the condition affected individuals’ ability to downplay its seriousness. The Importance o f Prevalence Information. In a survey of college students, those who had experienced a particular illness tended to rate it as being more common or prevalent than students who had not experienced the illness (Jemmott et al., 1988). These perceptions were related to perceived seriousness of a condition. That is, those who believed that the condition was common also tended to rate it as less serious (Croyle & Sande, 1988). Jemmott et al. (1986) directly manipulated how prevalent subjects believed the enzyme deficiency condition to be among subjects in a given session. Sub­ jects were told either that one in five of them had tested positive or that four in five of them had tested positive. When the deficiency was seen as rare (one in five), subjects rated it as more serious than when it was seen as common (four in five). In another study, Ditto and Jemmott (1989) manipulated the prevalence of the condition, as well as whether subjects believed it to be a positive (healthful) or a negative health characteristic. In the positive characteristic condition, the defi­ ciency was appraised as even more positive when it was thought to be rare than when it was thought to be common. For those who believed it to be a negative health characteristic, the deficiency was rated as more negative when it was rare than when it was common. The way in which the rarity of a condition affects its appraisal has been called a “scarcity heuristic” (Croyle & Ditto, 1990; Croyle & Jemmott, 1991; Ditto & Jemmott, 1989), in which perceptions of rarity lead to evaluative extremity. If a condition is perceived as good, rarity increases its perceived positivity; if the condition is perceived as bad, rarity increases its perceived negativity. Information regarding the experience and responses of others also may affect individuals’ responses to treatment and their recovery (Sanders, 1982; Sanders

270

CLARK

& Kardinal, 1977). The next section focuses on research examining social com­ parison processes relevant to recovery and adjustment to illness. Recovery and Adjustment to Illness

Research on social comparisons among individuals who suffer from various ill­ nesses has suggested that the major tenants of the original theory may not always hold true (Taylor, Buunk, & Aspinwall, 1990; Wood, 1989). In particular, in­ dividuals do not seem to be motivated by accuracy, but in many cases are moti­ vated to feel better about themselves by comparing themselves to others who are worse off (Taylor, Wood, & Lichtman, 1983; Wills, 1981, 1983). These “down­ ward comparisons” serve the function of self-enhancement, rather than selfevaluation. Self-Enhancement Through Downward Social Comparisons. There is evidence of spontaneous and relatively frequent downward comparisons occurring across many illness and health-relevant conditions. Specifically, downward social com­ parison has been shown to occur among individuals who experience spinal cord injury (Schulz & Decker, 1985), cancer (Taylor et al., 1983; Wood, Taylor, & Lichtman, 1985), rheumatoid arthritis (Affleck, Tennen, Pfeiffer, & Fifield, 1987, 1988; Affleck, Tennen, Pfeiffer, Fifield, & Rowe, 1987; DeVellis et al., 1990), infertility (Affleck & Tennen, 1991), aging (Suls, Marco, & Tobin, 1991), and parenting of a medically fragile infant (Affleck et al., 1987). Individuals choose to compare themselves with worse-off others, both in terms of their physical situation and their coping abilities and resources (Affleck & Ten­ nen, 1991; Collins, Dakof, & Taylor, 1988; Taylor, Falke, Shoptaw, & Licht­ man, 1986). Although making such comparisons was not related to clinician ratings of adjustment improvement in arthritis sufferers (DeVellis et al., 1990), such com­ parisons were more likely to be made by well-adjusted cancer patients (Taylor et al., 1983; Wood et al., 1985). In addition, Affleck et al. (1987) reported that the downward comparisons that parents made while their infant was in a neona­ tal intensive care unit were not related to self-reported mood advantages or nurse­ rated recovery. However, these social comparisons appeared to stay with the par­ ents in the form of recurring thoughts for 6 months, at which time they were positively correlated with self-reported positive mood, parental competency, and decreases in depression. Individuals'Reactions to the Social Environment. The previous findings sug­ gest that individuals may seek to compare themselves with others who are having a more difficult time than they are, either through the experience of more severe situations or because they are worse off in terms of coping abilities and resources. However, at times, individuals’ social environments may not cooperate with the search for worse off others (Wood, 1989). For example, individuals may be

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

271

socially isolated due to inhibitions of self-disclosure or a lack of knowledge of others who share similar conditions (Taylor & Lobel, 1989). Wood (1989) argued that, in such cases (and even when actual worse off others are available), individuals may creatively construct comparisons so as to be self­ enhancing. For example, individuals select specific dimensions on which they may be better off such as having more money or greater social support (Affleck et al., 1987; Taylor et al., 1983; Wood, 1989). Individuals deliberately may select unfamiliar others, in whom they can distort more easily the others’ situations to make themselves look better (Collins et al., 1992). In the absence of actual others, individuals may create a hypothetical compar­ ison person or manufacture group norms that make themselves look good (Af­ fleck, Tennen, & Rowe, 1988; Taylor et al., 1983; Wood, 1989). Individuals may choose nonsocial comparisons as well. For example, individuals think about how things might be worse, in essence comparing their situation with a hypothet­ ical worse world (Taylor et al., 1983). Or individuals may choose to make tem­ poral comparisons, in which they compare their present situations with their pasts (Albert, 1977; Masters & Keil, 1987; Perloff & Fetzer, 1986; Suls et al., 1991; Wood, 1989). Suls et al. studied both social and temporal comparisons among the elderly. They found that individuals showed favorable effects of downward social comparisons, but negative effects of temporal comparisons. Presumably, their pasts represented times of fewer health problems. The social environment may be uncooperative in a second way. At times, in­ dividuals may have forced comparisons thrust upon them (Wood, 1989). These comparisons may not always be favorable to the individual. Sources for such com­ parisons may include known others, the media (Taylor & Lobel, 1989), or sto­ ries of similarly suffering others provided by well-meaning friends and family (Affleck & Tennen, 1991; Taylor, Aspinwall, Dakof, & Reardon, 1988; Taylor, Collins, Skokan, & Aspinwall, 1989). Such comparisons are particularly painful when the dimensions of comparison is self-relevant (e.g., when the other is a competitor or the comparison involves some domain of particular importance to the individual; Tesser & Campbell, 1982; Tesser, Millar, & Moore, 1988). At times, the unavoidable comparison other may be someone to whom individuals are close, leading to envy and a need to create social distance from this person to avoid the unfavorable comparison and the feelings that it generates (Salovey & Rodin, 1984; Tesser, 1980; Wood, 1989). When faced with an unfavorable comparison, the individual may distort or highlight surrounding dimensions to make the actual social comparison irrele­ vant (Wood, 1989). For example, if a cancer patient learns that another patient already has returned to work activities after a similar surgery, he or she might point out that this individual is much younger, thus invalidating the comparison. Or the individual might choose to make additional “compensatory comparisons,” in which he or she restores his or her self-esteem through enhancing compari­ sons on other dimensions (Masters & Keil, 1987). For example, this cancer

272

CLARK

patient might feel that he or she is coping better with his or her family’s distress than his or her comparison other. In addition, it is possible that individuals attempt mentally to avoid the unflat­ tering comparison through thought suppression. Although this has not been test­ ed empirically, based on work by Wegner (1992; see also Cioffi & Holloway, 1993), we might expect that these individuals experience recurrent intrusive thoughts in the form of a “rebound” effect of the suppressed comparison. Such ruminations of a negative comparison may lead to poorer adjustment for individu­ als. Finally, being confronted with the progress or situation of better-off others need not have negative consequences. Indeed, Taylor and Lobel (1989) suggest­ ed that individuals may even seek out better-off others for purposes other than social comparisons. Making Downward Comparisons, but Upward Contacts. Individuals with cancer report that they react positively to hearing information about better-off others (Taylor et al., 1988). These others provide examples of the potential to cope well to their adverse circumstances, and they serve the function of role models (Taylor & Dakof, 1988). These role models can provide information important for problem-solving coping. They also may create feelings of inspiration and hope in individuals that motivate them to persevere in the face of their health problems (Taylor & Lobel, 1989). Upward and downward comparisons each have the potential to affect an in­ dividual positively or negatively (Buunk, Collins, Taylor, VanYperen, & Dakof, 1990; Taylor & Lobel, 1989). Seeing someone who is better off can make one feel either inferior or inspired. Encountering someone who is worse off can be self-enhancing, or it may point out a possible worse future awaiting the individu­ al. Buunk et al. (1990) found that cancer patients low in self-esteem were more likely to see the negative perspective in each type of comparison than were pa­ tients high in self-esteem (see also Wheeler & Miyake, 1992, for similar find­ ings with healthy individuals). In addition, the degree to which individuals perceived that they had some control over their prognoses and their symptoms determined their ability to gain advantage from downward comparisons. Those who perceived that they had little control showed more of a tendency to suffer as a result of making downward social comparisons. Theoretical Advances in Understanding the Process o f Social Compari­ sons. Social comparison information may be extracted from one’s social environ­ ment in the form of context (see Fig. 5.1). In addition, the preceding discussion of research suggested that comparisons may be generated selectively by the per­ ceiver, based on a combination of the stimulus and stored knowledge. Finally, in some cases, selectively retrieved knowledge is used to generate hypothetical others and manufacture norms.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

273

Recent review papers have pointed to the vast amount of cognitive flexibility inherent in the types of comparisons individuals experiencing health problems use (Taylor & Lobel, 1989; Wood, 1989). In making downward comparisons, individuals select a particular target as well as a dimension on which they appear advantaged. They also may choose unfamiliar others so that they can presume more easily that this target is worse off than themselves. When no actual com­ parison person is available, individuals create one. Or, they create norms of ad­ justment and hypothetical worse worlds. Finally, they may choose to make temporal comparisons between their present situations and a bad time in their own pasts. Even when individuals are confronted with upward comparison targets, they still have options for ameliorating the negative implications of such a compari­ son. After the initial encounter, they might choose to physically avoid interact­ ing with such individuals, making further comparisons less likely. Once the comparison has been made, individuals might focus selectively on a surrounding dimension on which this person is different, thus rendering the comparison in­ valid. Or they could make compensatory comparisons in which they merely select another dimension on which they are superior to the target. Finally, individuals might choose to use the target individual for purposes other than self-evaluation. For example, they might see this target as a role model who can provide them access to information useful for coping, as well as inspire positive emotions such as anticipation and hope. Table 5.1 shows the responses available to individuals who are confronted with comparison-relevant information about better-off others and the potential consequences of these responses. TA B LE 5.1 Response Options to Confronting Comparison-Relevant Information A bout a Better-off Other

Response Options

Potential Immediate Outcome

Mentally avoid/suppress the comparison

Rebound of unpleasant com­ parison in the form of in­ trusive thinking Create social distance from better-off others Render comparison irrelevant

Behaviorally avoid the comparison Highlight differences on surrounding dimension Make compensatory com­ parisons Use better-off other as a “contact” (role model)

Repair self-esteem damage Gain problem-solving infor­ mation Inspiration and hope act as motivators

Potential Long-Term Consequences

Negative effects on adjustment

Negative effects on social sup­ port (social isolation) Avoid negative effects of com­ parison Avoid negative effects of com­ parison Positive effects on adjustment

274

CLARK

SUMMARY

This chapter has highlighted the cognitive processes important to symptom ap­ praisal, reactions to diagnosis, health behaviors, and long-term adjustment to ill­ ness. Research has indicated that individuals store illness-relevant information in the form of structured knowledge schemas that affect perceptions and behaviors. In addition, their perceptions and behavioral decisions are colored by stable af­ fective dispositions. These affective orientations determine their vulnerability to illness, as well as their ability to cope with medical problems. Individuals are portrayed both as effective users of relevant health-related in­ formation and as motivated perceivers of themselves and others (Kunda, 1990). For example, individuals appear to be motivated to downplay the seriousness of their health conditions and to use constructive interpretative processes to see them­ selves as better off than others in similar states. At the same time, individuals are capable of taking advantage of information that will help them deal with health problems, in the form of a nurse describing sensations to expect or from betteroff role models. This chapter outlined five major areas in which health psychology research has addressed cognitive processes. Naturally, many theoretical perspectives were not represented. These include attributional processes in health (Michela & Wood, 1986), decision theory (see Linville et al., 1993), and personal control (Rodin & Salovey, 1989). In addition, many of the content areas of health psychology were not represented, including work on health attitudes, persuasion and behavior change, doctor-patient interaction, compliance, and the role that cognition plays in the relationships among stress, depression, and illness. Potential Theoretical Integrations in Two Domains in Health Psychology

Up to this point, domains of research in health psychology using a social cogni­ tion perspective have been discussed in relative isolation. I next present potential integrations of some of this work in two areas: health-related memory and long­ term adjustment to illness. These integrations represent established relationships that have been discussed in this chapter, as well as potential hypotheses that await empirical investigation. Recall o f Somatic Experience. Figure 5.2 shows one potential integration of the work relevant to the recall of somatic experience. The solid lines represent relationships that have been established in the literature; the dotted lines are hypotheses that have not yet been tested. In general, there are three paths of rela­ tionships among factors proposed to affect recall. The first of these is the subjec­ tive representation path. Put simply, subjective representations of illness engender confirmatory search strategies that lead to selective recall of illness confirming

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

Representations

Confirmatory

of Illness

Search Strategy

Focus on Sensation Information

Objective Features

Fewer Negative

Encoded (blocks

Aspects Encoded

275

--

threat value)

Focus on Sensation

Focus on Sensation

Sensation

Mood Congruent Recall Processes

Focus on Focus on Sensation Focus on Sensation

FIG. 5.2. An integration of the factors and processes determining recall of somatic experience.

symptoms (Croyle & Sande, 1988; Baumann et al., 1989; Bishop, et al., 1987). The second path, which is more speculative, draws on the work of Leventhal and his colleagues in investigating the impact of sensation information on the experience of physical sensations. Although recall of the experience of physical symptoms or pain has not been specifically studied in this line of work, one might draw some conclusions about how, for example, a focus on the sensations of a medical procedure might affect one's later recall of that experience. Sensation information leads the individual to create a schema for the experience that highlights the objective features of the sensation while blocking an interpretation of the experience as emotionally threatening (Leventhal, 1982; Leventhal & Johnson, 1983). This results in lowered emotional distress at the time of the experience. Both the focus on the objective elements and the lowered distress over the experience should lead to fewer negative aspects of the experience being encoded, and therefore later able to be recalled. The third path concerns the mood-congruent research on recall of healthrelevant experiences. The work of Salovey (Salovey & Birnbaum, 1989; Salovey et aI., 1991) and others (Croyle & Uretsky, 1987) showed that mood at encoding and at retrieval can affect one's recall of health-relevant experience. In addition, Eich et al. (1990) showed that pain at retrieval, if it increased one's negative mood, affected one's recall of earlier experiences. Although it has not been studied specifically, it is plausible that pain experienced at encoding could affect one's mood at encoding, thereby affecting mood-congruent recall in much the same fashion as pain experienced at the time of retrieval. In their review of sensation information studies, Taylor and Clark (1986) reported that sensation

276

CLARK

information reliably improved patients' postoperative mood and lowered the need for postoperative pain medication. In this way, sensation information could influence the mood-congruent processes involved in health-related memory. Finally, although again not yet studied explicitly, the role of stable personality characteristics, such as optimism and negative affectivity, most certainly affect an individual's mood at both encoding and retrieval in ways that will be likely to facilitate mood-congruent recall. For example, the stable disposition of someone high in negative affectivity is likely to ensure that this person is in a negative mood at both encoding and retrieval periods, thereby facilitating mood-congruent recall of pain or other potentially unpleasant sensations. While several of these relationships await empirical study, it is useful to think about how the processes described in this chapter might interrelate and affect the cognitive processes involved in memory. For one thing, such an integration can lead to new theorizing and hypotheses. In the next attempt at integration, I examine the processes involved in adjustment to long-term illness.

Adjustment to Long-Term Illness. Figure 5.3 outlines some of the possible relationships among factors described in this chapter that have been linked to longterm adjustment to illness. Once again, solid lines have been used to refer to established relationships or those previously discussed theoretically, and dotted lines to new hypotheses not yet tested. Although this particular integration does not draw on all the social cognitive health findings relevant to long-term adjustment, it does draw on several concepts and processes discussed in this chapter. Optimism



Expect

Cynical

Ability to

Hostility Self-esteem

Gain Benefits

Self­

from Social

Enhancement

Comparisons

Psychological

Enhanced

Control

Coping

Adjustment to Long-Term Illness

Perseverate

the Best

Confronting

^

Planning

FIG. 5.3. An integration of the factors and processes determining long-term adjustment to illness.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

Til

In the proposed model, there are three major paths to adjustment benefits: through optimism, through confronting (rather than avoidance), and through the benefits of both upward and downward social comparisons. In the first path, op­ timistic individuals are seen as gaining adjustment benefits primarily through their expectations that things will turn out for the best. These expectations facilitate planning and perseverance, which eventually lead to greater success in adjusting to illness (Scheier & Carver, 1985, 1992). The second path, suggested by the research in attentional processes in longrange adjustment, offers an explanation for why individuals who confront, rather than avoid, their illnesses show long-range benefits (Suls & Fletcher, 1985). Spe­ cifically, directing their attention toward the illnesses can lead to increased feel­ ings of psychological control that enable individuals to make use of coping strategies that directly affect their adjustment (Johnson, 1984; Leventhal & John­ son, 1983; Suls & Fletcher, 1985). At the same time, such individuals may benefit from an orientation toward planning for future obstacles and recovery-related goals that individuals who practice avoidance cannot (Johnson, 1984; Suls & Fletcher, 1985). The third pathway, suggested by research on social comparison benefits, speci­ fies benefits for coping that favorable comparisons can provide. Specifically, downward comparisons to worse-off others provide self-enhancement that may combat some deleterious effects of chronic illness, such as depression. Upward comparisons may provide inspiration that increases individuals’ perseverance, as well as new coping strategies that lead to enhanced adjustment (Kulick & Mah­ ler, 1987b; Taylor & Lobel, 1989). Two antecedents to individuals’ ability to make use of these positive consequences of social comparison are initial self-esteem and feelings of control (Buunk et al., 1990). Specifically, those high in self-esteem are more able to make use of both downward and upward comparisons. Individuals with little perceived control are more likely to see downward comparisons as threatening (e.g., “this is the future that awaits me”). Although it has not been studied empirically up to this point, it is interesting to think about ways in which cynical hostility might affect individuals’ ability to gain benefits from social comparisons. Specifically, the competitive and distrustful nature of cynically hostile people may make it unlikely that they are open to see­ ing those who are better off as potential role models. In addition, their propensi­ ty to respond to social situations with anger suggests that they might create interpersonal conflict and undermine social support, and hence antagonize potential comparison others (Hardy & Smith, 1988; Musante et al., 1989; Smith & Chris­ tianson, 1992). One also might predict that cynically hostile individuals are bet­ ter practiced at seeing others in a disadvantaged or negative light, thereby giving them a facilitation at downward comparisons. Although interesting, these hypotheses are speculative and await empirical investigation. These two diagrams show examples of the ways in which research in health psychology and cognitive processes might be integrated. Such integrations can

278

CLARK

tie together important literatures and, at the same time, generate new research hypotheses for the direction of future work. Hopefully the work discussed in this chapter has made clear the many ways in which health psychology has been af­ fected by work in social cognition and cognitive psychology. In ending this chapter, I point out the reasons why health psychology provides an excellent domain in which to conduct social cognition research. Health Psychology's Contributions to Social Cognition

Critics of social cognition research have argued that much work in this area is artificial, entailing a focus that has, in the past, excluded relevant social varia­ bles such as affect and interpersonal relationships. In addition, critics have com­ plained that social cognition often fails to demonstrate effects on “real” socially relevant behaviors (Martin & Clark, 1990). The work in health psychology report­ ed in this chapter has highlighted the need for more work on antecedent factors in information processing, as well as the importance of surrounding context and the ways in which response outcomes can feedback into the processing of subse­ quent stimuli. There are three clear advantages to studying social cognition processes in health psychology. First, health psychology concerns real-world phenomena, and hence provides a meaningful context in which to study the cognitions and behaviors of individuals. Indeed, such contexts often have life or death consequences for the individual, such as the experience of coronary disease, cancer, or AIDS (Folkman, in press; Linville et al., 1993). This is in keeping with social psychology’s concerns with relevant social issues and the recent efforts of social cognition researchers to examine real-world behavior and concerns. Second, health psychology focuses on topics that recently have presented challenges for the social cognition researcher. These include emotions, motiva­ tion, goals, and social relationships. Health psychology questions often revolve around the relationships among cognition, affect, and behavior. In fact, in health psychology, these concepts often are not easily separable. An example is the con­ text of AIDS, where individuals’ reactions to risk factor information may affect subsequent health-relevant behaviors involving others. For health psychologists, the primary concern of research endeavors is often the prediction and understand­ ing of actual behavior. Therefore, it affords the social cognition researcher an excellent context in which to develop and test theories that lead to the under­ standing of the cognition-behavior link. Finally, health psychology generally subscribes to a systems approach in the form of the biopsychosocial model of health and illness (Schwartz, 1982; Tay­ lor, 1991). This model values the interchange among social, psychological, and biological levels of human functioning as they contribute to health and illness. For example, studies of lifestyle change or adherence to complex medical regi­ mens may involve an understanding of the impact of physiological risk factors,

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

27 9

psychological reactions, and social values or practices as they impact on an in­ dividual’s behavior. Therefore, health psychology offers an arena in which to inte­ grate different levels of analysis, in particular, social cognition findings with more molecular (physiological) and molar (social and sociological) levels of analysis.

CONCLUSIONS

As in other work in social cognition, health psychology acknowledges the im­ portance of understanding cognitive processes to explain the phenomenology of individuals. There is strong evidence that such processes predict individuals’ likeli­ hood of becoming ill, how they respond to physical symptoms, and the ease of their recovery from illnesses. This work highlighted the flexibility inherent in the cognitive system, as well as the impact of affect, preexisting knowledge, and coping strategies. Perhaps no other field makes it more salient how profoundly day-to-day cognitive responses to incoming stimuli can affect people.

ACKNOWLEDGMENTS

The contribution made by this chapter has been enhanced by the help of several people. My knowledge and perspective on health and social cognition have been fostered in graduate school and beyond by my advisers, Shelley Taylor and Christine Dunkel-Schetter, and by Patricia Linville. I wish to thank Delia Cioffi and Peter Salovey for their much-respected insights and criticisms. I am also grate­ ful to Robert Wyer and Thom Srull, who, as editors, improved the specific con­ tents. Finally, I am indebted to several health psychology researchers who answered my request for their latest work.

REFERENCES Affleck, G., & Tennen, H. (1991). Social comparison and coping with major medical problems. In J. Suls & T. A. Willis (Eds.), Social comparison: Contemporary theory and research (pp. 369-393). Hillsdale, NJ: Lawrence Erlbaum Associates. Affleck, G., Tennen, H., Pfeiffer, C ., & Fifield, J. (1987). Appraisals of control and predictability in adapting to a chronic disease. Journal o f Personality and Social Psychology, 53, 273-279. Affleck, G., Tennen, H., Pfeiffer, C., & Fifield, J. (1988). Social comparisons in rheumatoid arthri­ tis: Accuracy and adaptational significance. Journal o f Social and Clinical Psychology, 6, 219-234. Affleck, G., Tennen, H., Pfeiffer, C., Fifield, J., & Rowe, J. (1987). Downward comparison and coping with serious medical problems. American Journal o f Orthopsychiatry, 57, 570-578. Affleck, G., Tennen, H., & Rowe, J. (1988). Adaptational features of mothers’ risk and prevention appraisals after the birth of high risk infants. American Journal o f Mental Retardation, 92, 360-368. Albert, S. (1977). Temporal comparison theory. Psychological Review, 84, 485-503.

280

CLARK

Allred, K. D., & Smith, T. W. (1989). The hardy personality: Cognitive and physiological responses to evaluative threat. Journal o f Personality and Social Psychology, 56, 257-266. Allred, K. D., & Smith, T. W. (1991). Social cognition in cynical hostility. Cognitive Therapy and Research, 15, 399-412. American Psychiatric Association. (1980). Diagnostic and statistical manual o f mental disorders (3rd ed.). Washington, DC: Author. Anderson, J. R., & Bower, G. H. (1979). Human associative memory. Hillsdale, NJ: Lawrence Erl­ baum Associates. Antoni, M. H., & Goodkin, K. (1988). Host moderator variables in the promotion of cervical neoplasis: Vol 1. Personality facets. Journal o f Psychosomatic Research, 32, 327-338. Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and social perception: The influence of trait information presented outside of conscious awareness of impression forma­ tion. Journal o f Personality and Social Psychology, 43, 437-449. Baumann, L. J., Cameron, L. D., Zimmerman, R. S., & Leventhal, H. (1989). Illness representa­ tions and matching labels with symptoms. Health Psychology, 8, 449-469. Bishop, G. D. (1987). Lay conceptions of physical symptoms. Journal o f A pplied Social Psycholo­ gy, 17, 127-146. Bishop, G. D. (1991). Understanding the understanding of illness. In J. A. Skelton & R. T. Croyle (Eds.), Mental representation in health and illness (pp. 32-59). New York: Springer-Verlag. Bishop, G. D., Alva, A. L., Cantu, T ., & Rittman, T. K. (1991). Responses to persons with AIDS: Fear of contagion or stigma? Journal o f Applied Social Psychology, 21, 1877-1888. Bishop, G. D., Briede, C., Cavazos, L., Grotzinger, R., & McMahon, S. (1987). Processing illness information: The role of disease prototypes. Basic and Applied Social Psychology, 8, 21-44. Bishop, G. D., & Converse, S. A. (1986). Illness representations: A prototype approach. Health Psychology, 5, 95-114. Blaney, N. T., Morgan, R. D., Feaster, D., Millon, C., Szapocznick, J., & Eisdorfer, C. (1991). Cynical hostility: A risk factor in HIV-1 infection? Journal o f A pplied Social Psychology, 21, 668-695. Bower, G. H. (1981). Emotional mood and memory. American Psychologist, 36, 129-148. Bower, G. H., & Meyer, J. D. (1989). In search of mood dependent retrieval. Journal o f Social Behavior and Personality, 4, 121-156. Brett, J. F., Brief, A. P., Burke, M. J., George, J. M ., & Webster, J. (1990). Negative affectivity and the reporting of stressful life events. Health Psychology, 9, 57-68. Brewer, W. F., & Nakamura, G. V. (1984). The nature and functions of schemas. In R. S. Wyer, Jr., & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 1, pp. 119-160). Hillsdale, NJ: Lawrence Erlbaum Associates. Burbach, D. J., & Peterson, L. (1986). Children’s concepts of physical illness: A review and critique of the cognitive-developmental literature. Health Psychology, 5, 307-325. Buunk, B. P., Collins, R. L., Taylor, S. E ., VanYperen, N. W ., & Dakof, G. A. (1990). The affec­ tive consequences of social comparison: Either direction has its ups and downs. Journal o f Per­ sonality and Social Psychology, 59, 1238-1249. Cacioppo, J. T., Andersen, B. L ., Tumquist, D. C., & Tassinary, L. G. (1989). Psychophysiological comparison theory: On the experience, description, and assessment of signs and symptoms. Patient Education and Counseling, 13, 257-270. Carlston, D. E., & Skowronski, J. J. (1986). Trait memory and behavior memory: The effects of alternative pathways on impression judgment response times. Journal o f Personality and Social Psychology, 50, 5-13. Carver, C. S., Scheier, M. F., & Pozo, C. (1992). Conceptualizing the process of coping with health problems. In H. S. Friedman (Ed.), Hostility, coping, and health (pp. 167-187). Washington, DC: American Psychological Association. Christianson, A. J., & Smith, T. W. (1991). Cynical hostility and cardiovascular reactivity during self-disclosure. Manuscript submitted for publication.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

281

Cioffi, D. (1991a). Beyond attentional strategies: A cognitive-perceptual model of somatic interpre­ tation. Psychological Bulletin, 109, 25-41. Cioffi, D. (1991b). Asymmetry of doubt in medical self-diagnosis: The ambiguity of “uncertain well­ ness.” Journal o f Personality and Social Psychology, 61, 969-980. Cioffi, D. (1993). Sensate body, directive mind: Physical sensations and mental control. In D. M. Wegner & J. W. Pennebaker (Eds.), Handbook o f mental control (pp. 410-442). Englewood Cliffs, NJ: Prentice-Hall. Cioffi, D. (in press). When good news is bad news: Medical wellness as a non-event in undergradu­ ates. Health Psychology. Cioffi, D ., & Holloway, J. (1993). The delayed costs of suppressed pain. Journal o f Personality and Social Psychology, 64, 274-282. Clark, L. F., Garstka, T ., Bless, H ., & Shah, R. (1992). M ood as information effects in se lf ap­ praisal o f coping style. Unpublished manuscript. Clark, L. F., Henry, S. M ., & Taylor, D. M. (1991). Cognitive examination of motivation for child­ bearing as a factor in adjustment to infertility. In A. L. Stanton & C. Dunkel-Schetter (Eds.), In­ fertility: Perspectives from stress and coping research (pp. 157-180). New York: Plenum. Cohen, F., & Lazarus, R. S. (1973). Active coping processes, coping processes, coping disposi­ tions, and recovery from surgery. Psychosomatic Medicine, 35, 375-389. Cohen, S., & Williamson, G. M. (1991). Stress and infectious disease in humans. Psychological Bulletin, 109, 5-24. Collins, A. M ., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psy­ chological Review, 83, 407-428. Collins, R. F., Dakof, G., & Taylor, S. E. (1992). Social comparison and adjustment to a threaten­ ing event. Manuscript submitted for publication. Croyle, R. T. (1990). Biased appraisal of high blood pressure. Preventive Medicine, 19, 40-44. Croyle, R. T. (1992). Appraisal of health threats: Cognition, motivation, and social comparison. Cognitive Therapy and Research, 16, 165-182. Croyle, R. T ., & Barger, S. D. (1993). Illness cognition. In N. S. Maes, H. Leventhal, & M. John­ ston (Eds.), International review o f health psychology (Vol. 2, pp. 29-49). Chichester: John Wiley & Sons. Croyle, R. T ., & Ditto, P. H. (1990). Illness cognition and behavior: An experimental approach. Journal o f Behavioral Medicine, 13, 31-52. Croyle, R. T ., & Hunt, J. R. (1991). Coping with health threat: Social influence processes in reac­ tions to medical test results. Journal o f Personality and Social Psychology, 60, 382-389. Croyle, R. T ., & Jemmott III, J. B. (1991). Psychological reactions to risk factor testing. In J. A. Skelton & R. T. Croyle (Eds.), Mental representation in health and illness (pp. 85-107). New York: Springer-Verlag. Croyle, R. T ., & Loftus, E. F. (in press). Recollection in the kingdom of AIDS. In D. G. Ostrow & R. Kessler (Eds.), AIDS Mental Health Survey Research M ethodology. New York: Plenum. Croyle, R. T ., & Loftus, E. F. (1992). Improving episodic memory performance of survey respond­ ents. In J. M. Tanur (Ed.), Questions about questions: Inquiries into the cognitive bases o f surveys (pp. 95-101). New York: Russell Sage. Croyle, R. T., Sun, Y., & Louie, D. (in press). Psychological minimization of cholesterol test results: Moderators of appraisal in college students and community residents. Health Psychology. Croyle, R. T ., & Sande, G. N. (1988). Denial and confirmatory search: Paradoxical consequences of medical diagnosis. Journal o f Applied Social Psychology, 18, 473-490. Croyle, R. T ., & Uretsky, M. B. (1987). Effects of mood on self-appraisal of health status. Health Psychology, 6, 239-253. Croyle, R. T ., & Williams, K. D. (1991). Reactions to medical diagnosis: The role of illness stereo­ types. Basic and Applied Social Psychology, 12, 227-241. Dembroski, T. M ., MacDougall, J. M ., Costa, P. T ., Jr., & Grandits, G. A. (1989). Components of hostility as predictors of sudden death and myocardial infarction in the Multiple Risk Factor Intervention Trial. Psychosomatic Medicine, 51, 514-522.

282

CLARK

DeVellis, R., Holt, K., Renner, B., Blalock, S., Blanchard, I., Cook, H ., Klotz, M ., Mikow, V., & Harring, K. (1990). The relationship of social comparison to rheumatoid arthritis symptoms and affect. Basic and Applied Social Psychology, 11, 1-18. Ditto, P. H., & Jemmott, J. B. (1989). From rarity to evaluative extremity: Effects of prevalence information on the evaluation of positive and negative characteristics. Journal o f Personality and Social Psychology, 57, 16-26. Ditto, P. H., Jemmott, J. B., & Darley, J. M. (1988). Appraising the threat of illness: A mental representational approach. Health Psychology, 7, 183-200. Eich, E., Rachman, S., & Lopatka, C. (1990). Affect, pain, and autobiographical memory. Journal o f Abnormal Psychology, 99, 174-178. Fazio, R. H. (1987). Self-perception theory: A current perspective. In M. P. Zanna, J. M. Olson, & C. P. Herman (Eds.), Social influence: The Ontario Symposium (Vol. 5, pp. 129-149). Hills­ dale, NJ: Lawrence Erlbaum Associates. Fazio, R. H., Sherman, S. J., & Herr, P. M. (1982). The feature-positive effect in the self-perception process: Does not doing matter as much as doing? Journal o f Personality and Social Psychology, 42, 404-411. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117-140. Fish, T. A., & Rye, B. J. (1991). Attitudes towards a homosexual or heterosexual person with AIDS. Journal o f Applied Social Psychology, 21, 651-657. Fisher, R. P., & Quigley, K. L. (1992). Applying cognitive theory in public health investigation: Enhancing food recall with the cognitive interview. In J. Tanur (Ed.), Questions about questions: Inquiries into the cognitive bases o f surveys (pp. 154-169). New York: Russell Sage. Fiske, S. T. (1982). Schema-triggered affect: Applications to social perception. In M. S. Clark & S. T. Fiske (Eds.), Affect and cognition: The 17th Annual Carnegie Symposium on Cognition (pp. 55-78). Hillsdale, NJ: Lawrence Erlbaum Associates. Fiske, S. T., & Taylor, S. E. (1991). Social Cognition (2nd ed.). New York: McGraw-Hill. Folkman, S. (in press). Psychosocial effects of HIV infection. In L. Goldberger & S. Breznitz (Eds.), Handbook o f stress (2nd ed.). New York: Free Press. Forston, M. T., & Stanton, A. L. (1992). Self-discrepancy theory as a framework for understanding bulimic symptomatology and associated distress. Journal o f Social and Clinical Psychology, 11, 103-118. Friedman, H. S. (1993). Interpersonal expectations and the maintenance of health. In P. D. Blanck (Ed.), Interpersonal expectations: Theory, research, and applications (pp. 179-193). New York: Cambridge University Press. Friedman, H. S., & Booth-Kewley, S. (1987). The “disease-prone” personality: A meta-analytic view of the construct. American Psychologist, 42, 539-555. Garrity, T. F., McGill, A., Becker, M ., Blanchard, E., Crews, J., Cullen, J., Hackett, T., Taylor, J., & Valins, S. (1976). Report of the task group on cardiac rehabilitation. In S. M. Weiss (Ed.), Proceedings o f the National Heart and Lung Institute working conference on health behavior (DHEW Publication No. 76-868). Washington, DC: U.S. Government Printing Office. Gonder-Frederick, L. A., Cox, D. J., Bobbitt, S. A., & Pennebaker, J. N. (1986). Blood glucose symptom beliefs of diabetic patients: Accuracy and implications. Health Psychology, 5, 327-341. Hardy, J. D., & Smith, T. W. (1988). Cynical hostility and vulnerability to disease: Social support, life stress, and physiological response to conflict. Health Psychology, 7, 447-459. Hastie, R. (1981). Schematic principles in human memory. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario Symposium (Vol. 1, pp. 39-88). Hillsdale, NJ: Lawrence Erlbaum Associates. Hastie, R., Park, B., & Weber, R. (1984). Social memory. In R. S. Wyer, Jr., & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 2, pp. 151-212). Hillsdale, NJ: Lawrence Erlbaum Associates. Heath, L., Acklin, M ., & Wiley, K. (1991). Cognitive heuristic and AIDS risk assessment among physicians. Journal o f A pplied Social Psychology, 21, 1859-1867.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

2 83

Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Review, 94, 319-340. Higgins, E. T ., & King, G. A. (1981). Accessibility of social constructs: Information-processing consequences of individual and contextual variability. In N. Cantor & J. F. Kihlstrom (Eds.), P er­ sonality cognition and social interaction (pp. 69-122). Hillsdale, NJ: Lawrence Erlbaum Associates. Higgins, E. T., Vookles, J., & Tykocinski, O. (1992). Self and health: How “patterns” of self-beliefs predict types of emotional and physical problems. Social Cognition, 10, 125-150. Hintzman, D. L. (1986). “Schema abstraction” in a multiple-trace memory model. Psychological Review, 93, 411-428. Janis, I. L. (1958). Psychological stress. New York: Wiley. Jemmott, J. B., Croyle, R. T., & Ditto, P. H. (1988). Common sense epidemiology: Self-based judg­ ments from laypersons and physicians. Health Psychology, 7, 55-73. Jemmott, J. B., Ditto, P. H., & Croyle, R. T. (1986). Judging health status: Effects of perceived prevalence and personal relevance. Journal o f Personality and Social Psychology, 50, 899-905. Johnson, J. E. (1975). Stress reduction through sensation information. In I. G. Sarason & C. D. Spielberger (Eds.), Stress and anxiety (Vol. 2, pp. 361-378). Washington, DC: Hemisphere. Johnson, J. E. (1984). Psychological interventions and coping with surgery. In A. Baum, S. E. Tay­ lor, & J. E. Snyder (Eds.), Handbook o f psychology and health (Vol. 4, pp. 167-187). Hillsdale, NJ: Lawrence Erlbaum Associates. Johnson, J. E., Lauver, D. R., & Nail, L. M. (1989). Process of coping with radiation therapy. Journal o f Consulting and Clinical Psychology, 57, 358-364. Johnson, J. E., & Leventhal, H. (1974). Effects of accurate expectations and behavioral instructions on reactions during a noxious medical examination. Journal o f Personality and Social Psychology, 29, 710-718. Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. American Psychologist, 39, 341-350. Kent, G. (1985). Memory for dental pain. Pain, 21, 187-194. Koskenvuo, M ., Kapiro, J., Rose, R. J., Kesnaiemi, A., Sarnaa, S., Heikkila, K., & Langivanio, H. (1988). Hostility as a risk factor for mortality and ischemic heart disease in men. Psychosomat­ ic Medicine, 50, 330-340. Krantz, D. S., & Deckel, A. W. (1983). Coping with coronary heart disease and stroke. In T. G. Burish & L. A. Bradley (Eds.), Coping with chronic disease: Research and applications (pp. 16-38). New York: Academic Press. Kulick, J. A., & Mahler, H. I. M. (1987a). Health status, perceptions of risk, and prevention in­ terest for health and nonhealth problems. Health Psychology, 6, 15-27. Kulick, J. A., & Mahler, H. I. M. (1987b). Effects of preoperative roommate assignment on preoper­ ative anxiety and recovery from coronary-bypass surgery. Health Psychology, 6, 525-543. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480-498. Lau, R. R., Bernard, T. M., & Hartman, K. A. (1989). Further explorations of common sense represen­ tations of common illnesses. Health Psychology, 8, 195-219. Lau, R. R., & Hartman, K. A. (1983). Common sense representations of illness. Health Psycholo­ gy, 2, 167-185. Lazarus, R. S. (1983). The costs and benefits of denial. In S. Breznitz (Ed.), Denial o f stress (pp. 1-30). New York: International Universities Press. Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer-Verlag. Lazarus, R. S., Kanner, A. D., & Folkman, S. (1980). Emotions: A cognitive-phenomenological analysis. In R. Plutchik & H. Kellerman (Eds.), Emotion: Theory, research and experience (pp. 189-217). New York: Academic Press. Leiker, M ., & Hailey, B. J. (1988). A link between hostility and disease: Poor health habits? Be­ havioral Medicine, 14, 129-133. Leventhal, H. (1982). The integration of emotion and cognition: A view from the perceptual motor theory of emotion. In M. S. Clark & S. T. Fiske (Eds.), Affect and Cognition (pp. 121-156). Hillsdale, NJ: Lawrence Erlbaum Associates.

284

CLARK

Leventhal, H. (1986). Symptom reporting: A focus on process. In S. McHugh & T. M. Vallis (Eds.), Illness Behavior: A multidisciplinary model (pp. 24-46). New York: Plenum. Leventhal, H., Brown, D., Shacham, S., & Engquist, C. (1979). Effects of preparatory information about sensations, threats of pain, and attention on cold pressor distress. Journal o f Personality and Social Psychology, 37, 688-714. Leventhal, H., & Diefenbach, M. (1991). The active side of illness cognition. In J. A. Skelton & R. T. Croyle (Eds.), Mental representation in health and illness (pp. 247-272). New York: SpringerVerlag. Leventhal, H., & Johnson, J. E. (1983). Laboratory and field experimentation: Development of the­ ory of self-regulation. In R. Lenard & P. Woodridge (Eds.), Behavioral science and nursing theory (pp. 189-262). St. Louis, MO: Mosby. Leventhal, H., Meyer, D., & Nerenz, D. (1980). The commonsense representation of illness danger. In S. Rachman (Ed.), M edical psychology (Vol. 2, pp. 7-30). New York: Pergamon. Leventhal, H., Nerenz, D. R., & Steele, D. J. (1984). Illness representations and coping with health threats. In A. Baum, S. E. Taylor, & J. E. Singer (Eds.), Handbook o f psychology and health (Vol. 4, pp. 219-252). Hillsdale, NJ: Lawrence Erlbaum Associates. Levine, M. N., Guyatt, G. H ., Gent, M ., De Pauw, S., Goodyear, M. D., Hryniuk, W. M ., Ar­ nold, A., Findlay, B., Skillings, J. R., Bramwell, V. H., Levin, L ., Bush, H., Abu-Zahra, H., & Kotalik, J. (1988). Quality of life in Stage II breast cancer: An instrument for clinical trials. Journal o f Clinical Oncology, 6, 1798-1810. Levy, S. M., Lee, J. K., Bagley, C., & Lippman, M. (1988). Survival hazards analysis in first recurrent breast cancer patients: Seven-year follow-up. Psychosom atic Medicine, 50, 520-528. Linville, P. W. (1987). Self-complexity as a cognitive buffer against stress-related depression and illness. Journal o f Personal and Social Psychology, 52, 663-676. Linville, P. W ., Fischer, G. W ., & Fischhoff, B. (1993). AIDS risk perceptions and decision biases. In J. B. Pryor & G. D. Reeder (Eds.), The social psychology o f infection (pp. 5-38). Hillsdale, NJ: Lawrence Erlbaum Associates. Lord, C. G., Lepper, M. R., & Preston, E. (1984). Considering the opposite: A corrective strategy for social judgment. Journal o f Personality and Social Psychology, 47, 1231-1243. Marato, J. J., Sheppard, J. A., & Pbert, L. A. (1990). D ispositional optimism as a predictor o f be­ havior changes among cardiac patients. Manuscript submitted for publication. Martin, L. L., & Clark, L. F. (1990). Social cognition: Exploring the mental processes involved in human social interaction. In M. W. Eysenck (Ed.), Cognitive psychology: An international review (pp. 266-310). Sussex, England: Wiley. Masters, J. C., & Keil, L. J. (1987). Generic comparison processes in human judgement and be­ havior. In J. C. Masters & W. P. Smith (Eds.), Social comparison, social justice, and relative deprivation: Theoretical, empirical, and policy perspectives (pp. 11-54). Hillsdale, NJ: Lawrence Erlbaum Associates. Matthews, K. A. (1988). CHD and Type A behaviors: Update on and alternative to the Booth-Kewley and Friedman quantitative review. Psychological Bulletin, 104, 373-380. Melzack, R., & Wall, P. D. (1982). The challenge o f pain. New York: Basic Books. Meyerowitz, B. E. (1983). Postmastectomy coping strategies and quality of life. Health Psychology, 2, 117-132. Meyerowitz, B. E ., & Chaiken, S. (1987). The effect of message framing on breast self-examination attitudes, intentions, and behavior. Journal o f Personality and Social Psychology, 52, 500-510. Michela, J. L., & Wood, J. V. (1986). Causal attributions in health and illness. Advances in cognitivebehavioral research (pp. 179-235). New York: Academic Press. Miller, S. M. (1979). Coping with impending stress: Psychophysiological and cognitive correlates of choice. Psychophysiology, 16, 575-581. Morgan, M. G., Fischhoff, B., Bostrom, A., Lave, L., & Atman, C. J. (in press). Risk communica­ tion (Prepared for reviews for American Scientist).

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

285

Mullen, B., & Suls, J. (1982). The effectiveness of attention and rejection as coping styles. Journal o f Psychosomatic Research, 26, 43-49. Musante, L., MacDougall, J. M ., Dembroski, T. M ., & Costa, P. T. (1989). Potential for hostility and dimensions of anger. Health Psychology, 8, 343-354. Namir, S., Wolcott, D. L., Fawzy, F. I., & Alumbaugh, M. J. (1987). Coping with AIDS: Psychol­ ogy and health implications. Journal o f Applied Social Psychology, 17, 309-328. Norvell, K. T ., Gaston-Johansson, F., & Fridh, G. (1987). Remembrance of labor pain: How valid are retrospective pain measurements? Pain, 31, 77-86. O’Sullivan, C. S., & Durso, F. T. (1984). Effect of schema-incongruent information on memory for stereotypical attributes. Journal o f Personality and Social Psychology, 47, 55-70. Pennebaker, J. W. (1982). The psychology o f physical symptoms. New York: Springer-Verlag. Perloff, L. S., & Fetzer, B. K. (1986). Self-other judgments and perceived vulnerability to victimi­ zation. Journal o f Personality and Social Psychology, 50, 502-510. Peterson, C., Seligman, M. E. P., & Vaillant, G. E. (1988). Pessimistic explanatory style is a risk factor for physical study. Journal o f Personality and Social Psychology, 55, 23-27. Prohaska, T. R., Keller, M. L., Leventhal, E. A., & Leventhal, H. (1987). Impact of symptoms and aging attribution on emotions and coping. Health Psychology, 6, 495-514. Pryor, J. B., & Reeder, G. D. (1993). Collective and individual representations of HIV/AIDS stig­ ma. In J. B. Pryor & G. D. Reeder (Eds.), The social psychology o f HIV infection (pp. 263-286). Hillsdale, NJ: Lawrence Erlbaum Associates. Pryor, J. B., Reeder, G. D., & McManus, J. (1991). Fear and loathing in the workplace: Reactions to AIDS-infected co-workers. Personality and Social Psychology Bulletin, 17, 133-139. Pyszczynski, T. A., & Greenberg, J. (1987). Self-regulatory perseveration and the depressive selffocusing style: A self-awareness theory of reactive depression. Psychological Bulletin, 102, 122-138. Robbins, A. S., Spence, J. T ., & Clark, H. (1991). Psychological determinants of health and perfor­ mance: The tangled web of positive and negative affectivity. Journal o f Personality and Social Psychology, 61, 755-765. Roche, P. A., & Gijsbers, K. (1986). A comparison of memory for induced ischemic pain and chronic rheumatoid pain. Pain, 25, 337-343. Rodin, J., & Salovey, P. (1989). Health psychology. Annual Review o f Psychology, 40, 533-579. Rumelhart, D. E., & Ortony, A. (1977). The representation of knowledge in memory. In R. C. An­ derson, R. J. Spiro, & W. E. Montague (Eds.), Schooling and the acquisition o f knowledge (pp. 99-136). Hillsdale, NJ: Lawrence Erlbaum Associates. Salovey, P., & Birnbaum, D. (1989). Influence of mood on health-relevant cognitions. Journal o f Personality and Social Psychology, 57, 539-551. Salovey, P., O’Leary, A., Stretton, M. S., Fiskin, S. A., & Drake, C. A. (1991). Influence of mood on judgments about health and illness. In J. P. Forgas (Ed.), Emotion and social judgments (pp. 241-262). Oxford, England: Pergamon. Salovey, P., & Rodin, J. (1984). Some antecedents and consequences of social-comparison jealousy. Journal o f Personality and Social Psychology, 47, 780-792. Salovey, P., Sieber, W. J., Jobe, J. B., & Willis, G. B. (in press). The recall of physical pain. In N. Schwarz & S. Sudman (Eds.), Autobiographical memory and the validity o f retrospective reports. New York: Springer-Verlag. Salovey, P., Sieber, W. J., Smith, A. F., Turk, D. C., Jobe, J. B., & Willis, G. B. (1992). Report­ ing chronic pain episodes on health surveys. Vital and Health Statistics, 6(6), 1-71. Sanders, G. S. (1982). Social comparison and perceptions of health and illness. In G. S. Sanders & J. Suls (Eds.), Social psychology o f health and illness (pp. 129-157). Hillsdale, NJ: Lawrence Erlbaum Associates. Sanders, J. B., & Kardinal, C. G. (1977). Adaptive coping mechanisms in adult acute leukemia pa­ tients in remission. Journal o f the American M edical Association, 238, 952-954.

286

CLARK

Scheier, M. F., & Carver, C. S. (1985). Optimism, coping, and health: Assessments and implica­ tions of generalized outcome expectancies. Health Psychology, 4 , 219-247. Scheier, M. F., & Carver, C. S. (1992). Effects of optimism on psychological and physical well­ being. Theoretical overview and empirical update. Cognitive Therapy and Research, 16, 201-228. Scheier, M. F., Matthews, K. A., Owens, J. F., Magovern, G. J., Sr., Lefebvre, R. C., Abbott, R. A., & Carver, C. S. (1989). Dispositional optimism and recovery from coronary artery bypass surgery: The beneficial effects on physical and psychological well-being. Journal o f Personality and Social Psychology, 57, 1024-1040. Schulz, R., & Decker, S. (1985). Long-term adjustment to physical ability: The role of social sup­ port, perceived control, and self-blame. Journal o f Personality and Social Psychology, 48, 1162-1172. Schwartz, G. E. (1982). Testing the biopsychosocial model: The ultimate challenge facing behavioral medicine? Journal o f Consulting and Clinical Psychology, 50, 1040-1053. Schwarz, N. (1990). Feelings as information: Informational and motivational functions of affective states. In E. T. Higgins & R. M. Sorrentino (Eds.), Handbook o f motivation and cognition: Foun­ dations o f social behavior (Vol. 2, pp. 527-561). New York: Guilford. Shapiro, D. E., Boggs, S. R., Melamed, B. G., & Graham-Pole, J. (1992). The effect of varied physician affect on recall, anxiety, and perceptions in women at risk for breast cancer: An ana­ logue study. Health Psychology, 11, 61-66. Skelton, J. A. (1987, May). Symptom suggestibility: Somatic imagery as an underlying process ? Present­ ed at the 48th Annual Convention of the Canadian Psychological Association, Vancouver, British Columbia. Skelton, J. A., & Croyle, R. T. (Eds.). (1991). Mental representation in health and illness. New York: Springer-Verlag. Skelton, J. A., & Pennebaker, J. W. (1990). The verbal system. In J. T. Cacioppo & L. G. Tassinary (Eds.), Principles o f psychophysiology: Physical, social, and inferential elements (pp. 631-657). New York: Cambridge University Press. Smith, E. E., & Medin, D. L. (1981). Categories and concepts. Cambridge, MA: Harvard Univer­ sity Press. Smith, E. R., & Zarate, M. A. (1992). Exemplar-based model of social judgment. Psychological Review, 99, 3-21. Smith, T. W ., & Allred, K. D. (1989). Blood pressure responses during social interaction in high and low cynically hostile males. Journal o f Behavioral Medicine, 12, 135-143. Smith, T. W ., & Christensen, A. J. (1992). Hostility, health, and social contexts. In H. S. Friedman (Ed.), Hostility, coping, and health (pp. 33-48). Washington, DC: American Psychological As­ sociation. Smith, T. W ., & Pope, M. K. (1990). Cynical hostility as a health risk: Current status and future directions. Journal o f Social Behavior and Personality, 5, 77-88. Smith, T. W ., Pope, M. K., Rhodewalt, F., & Poulton, J. L. (1989). Optimism, neuroticism, cop­ ing and symptom reports: An alternative interpretation of the Life Orientation Test. Journal o f Personality and Social Psychology, 56, 640-648. Strauman, T. J., Lemieux, A. M ., & Coe, C. L. (1993). Self-discrepancy and natural killer cell activity: Immunological consequences of negative self-evaluation. Journal o f Personality and So­ cial Psychology, 64, 1042-1052. Strauman, T. J., Vookles, J., Berenstein, V., Chaiken, S., & Higgins, E. T. (1991). Self-discrepancies and vulnerability and body dissatisfaction and disordered eating. Journal o f Personality and Social Psychology, 61, 946-956. Suls, J., Marco, C. A., & Tobin, S. (1991). The role of temporal comparison, social comparison and direct appraisal in the elderly’s self-evaluation of health. Journal o f Applied Social Psycholo­ gy, 21, 1125-1144. Suls, J., & Fletcher, B. (1985). The relative efficacy of avoidant and nonavoidant coping strategies: A meta-analysis. Health Psychology, 4, 249-288.

5.

SOCIAL COGNITION AND HEALTH PSYCHOLOGY

287

Suls, J., & Wan, C. K. (1982). Effects of sensory and procedural information on coping with stress­ ful medical procedures and pain: A meta-analyses. Journal o f Consulting and Clinical Psycholo­ gy, 57, 372-379. Taylor, S. E. (1991). Health psychology (2nd ed.). New York: McGraw-Hill. Taylor, S. E., Aspinwall, L. G., Dakof, G. A., & Reardon, K. (1988). Stress, storytelling, social comparison and social support: Victim’s reactions to stories o f similar victims. Manuscript in prepa­ ration. Taylor, S. E., Buunk, B. P., & Aspinwall, L. G. (1990). Social comparison, stress, and coping. Personality and Social Psychology Bulletin, 16, 74-89. Taylor, S. E., & Clark, L. F. (1986). Does information improve adjustment to noxious medical proce­ dures? In M. Sake & R. F. Kidd (Eds.), Advances in applied social psychology (Vol. 4, pp. 1-28). Hillsdale, NJ: Lawrence Erlbaum Associates. Taylor, S. E., Collins, R. L., Skokan, L. A., & Aspinwall, L. G. (1989). Maintaining positive illu­ sions in the face of negative information: Getting the facts without letting them get to you. Journal o f Social and Clinical Psychology, 8, 114-129. Taylor, S. E., & Dakof, G. A. (1988). Social support and the cancer patient. In S. Spacapan & S. Oskamp (Eds.), The social psychology o f health: The Claremont symposium on applied social p sy ­ chology (pp. 95-116). Newbury Park, CA: Sage. Taylor, S. E., Falke, R. L., Shoptaw, S. J., & Lichtman, R. R. (1986). Social support, support groups and the cancer patient. Journal o f Consulting and Clinical Psychology, 54, 608-615. Taylor, S. E., Kemeny, M. E., Aspinwall, C. G., Schneider, S. G., Rodriguez, R., & Herbert, M. (1992). Optimism, coping, psychological distress, and high-rise sexual behavior among men at risk for AIDS. Journal o f Personality and Social Psychology, 63, 460-473. Taylor, S. E., & Lobel, M. (1989). Social comparison activity under threat: Downward evaluation and upward contacts. Psychological Review, 96, 569-575. Taylor, S. E., Wood, J. V., & Lichtman, R. R. (1983). It could be worse: Selective evaluation as a response to victimization. Journal o f Social Issues, 39, 19-40. Tesser, A. (1980). Self-esteem maintenance in family dynamics. Journal o f Personality and Social Psychology, 39, 77-91. Tesser, A., & Campbell, J. (1982). Self-evaluation maintenance and the perception of friends and strangers. Journal o f Personality, 50, 261-279. Tesser, A., Millar, M ., & Moore, J. (1988). Some affective consequences of social comparison and reflection processes: The pain and pleasure of being close. Journal o f Personality and Social P sy­ chology, 54, 49-61. Thompson, S. C. (1981). Will it hurt less if I can control it? A complex answer to a simple question. Psychological Bulletin, 90, 89-101. Turk, D. C., Rudy, T. F., & Salovey, P. (1986). Implicit models of illness. Journal o f Behavioral Medicine, 9, 453-473. Tversky, A., Sattath, S., & Slovic, P. (1988). Contingent weighting in judgement and choice. P sy­ chological Review, 95, 371-384. Vallacher, R. R., & Wegner, D. M. (1985). A theory o f action identification. Hillsdale, NJ: Lawrence Erlbaum Associates. Vallacher, R. R., & Wegner, D. M. (1987). What do people think they’re doing? Action identifica­ tion and human behavior. Psychological Review, 94, 3-15. Van Komen, R. W ., & Redd, W. H. (1985). Personality factors associated with anticipatory nausea/vomiting in patients receiving cancer chemotherapy. Health Psychology, 4, 189-202. Wagener, J. J., & Taylor, S. E. (1986). What else could I have done? Health Psychology, 5, 481-496. Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress and distress: Exploring the cen­ tral role of negative affectivity. Psychological Review, 96, 233-253. Wegner, D. M. (1992). You can’t always think what you want: Problems in the suppression of un­ wanted thoughts. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 193-226). San Diego, CA: Academic Press.

288

CLARK

Weinstein, N. D. (1982). Unrealistic optimism about susceptibility to health problems. Journal o f Behavioral M edicine, 10, 481-500. Weinstein, N. D. (1983). Reducing unrealistic optimism about illness susceptibility. Health Psychol­ ogy, 2, 11-20. Wheeler, L., & Miyake, K. (1992). Social comparison in everyday life. Journal o f Personality and Social Psychology, 62, 760-773. Williams, R. B., Jr., Barefoot, J. C., & Shekelle, R. B. (1985). The health consequences of hostili­ ty. In M. A. Chesney & R. H. Rosemann (Eds.), Anger and hostility in cardiovascular and be­ havioral disorders (pp. 173-185). Washington, DC: Hemisphere. Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90, 245-271. Wills, T. A. (1983). Social comparison in coping and help-seeking. In B. M. DePaulo, A. Nadler, & J. D. Fisher (Eds.), New directions helping (Vol. 2, pp. 109-141). New York: Academic Press. Wood, J. V. (1989). Theory and research concerning social comparison of personal attributes. Psy­ chological Bulletin, 106, 231-248. Wood, J. V., Taylor, S. E., & Lichtman, R. R. (1985). Social comparison in adjustment to breast cancer. Journal o f Personality and Social Psychology, 49, 1169-1183. Wyer, R. S., Jr., & Carlston, D. E. (1979). Social cognition, inference, and attribution. Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., & Gordon, S. E. (1984). The recall of information about persons and groups. In R. S. Wyer, Jr., & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 2, pp. 73-150). Hills­ dale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., & Srull, T. K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., Srull, T. K., Gordon, S. E., & Hartwick, J. (1982). Effects of processing objec­ tives on the recall of prose material. Journal o f Personality and Social Psychology, 48, 533-549.

6

Social Cognition and Clinical Psychology: Anxiety, Depression, and the rocessing of Social Information P

John A. Edwards Ohio State University

Gifford Weary

Contents Cognitive Theories of Depression and Anxiety 292 Becks’ Schema Theory 292 294 The Helplessness-Hopelessness Model Self-Discrepancy Theory 295 Summary 296 Schematic Biases in Depression and Anxiety 296 Thought Content 297 Cognitive Processes in Depression 298 Cognitive Processes in Anxiety 301 Summary 302 Schematic Effects in the Perception of Others 302 Depression and the Perception of Others 304 Anxiety and the Perception of Others 312 Summary 315 Social-Cognitive Effects Related to Motivational Components of Depression and Anxiety 316 Control Motivation 317 Social Comparison Motives 321 Summary 322 Conclusion 323 Epilogue 325 Acknowledgments 326 References 326

As was the case in many areas of psychology, the 1970s saw a cognitive revolu­ tion in clinical psychology (Dember, 1974). During this time, a number of social learning and cognitive behavioral formulations concerned with the study of cog-

289

290

WEARY AND EDWARDS

nitive processes in the understanding of abnormal behavior appeared (Bandura, 1974, 1977; Beck, 1976; Foreyt & Rathjen, 1978; Mahoney, 1974). In 1976, a new journal, Cognitive Therapy and Research, devoted to empirical work on these topics, emerged. Most of the early cognitive work focused on the role of internal thoughts, expectations, self-statements, and feelings in the etiology, main­ tenance, and modification of numerous dysfunctional behaviors. More recent work at the interface of clinical and cognitive psychology has employed a more molecular level of analysis. Specifically, clinical scientists interested in the role of cogni­ tive processes in psychopathology and clinical practice have increasingly adopt­ ed ah information processing approach to clinical phenomena (Ingram, 1986). Regardless of the level of analysis, the current interest in cognition is so widespread that it has been said that “cognitive psychology has become mainstream clinical psychology” (Ingram, 1986, p. 3). The cognitive revolution within social psychology was occurring at about the same time as that within clinical psychology. Ostrom (1984) noted that social cognition was a popular topic of convention papers and edited volumes during the latter half of the 1970s and the early 1980s. Additionally, the journal Social Cognition appeared in 1982, and the first edition of the Handbook o f Social Cog­ nition was published in 1984. Today, work concerned with the perception, com­ prehension, memory, and judgments of the self and other people, as well as the myriad of social factors that can affect such cognitive processes, is a dominant theme in all areas of social psychology. Given these parallel trends within the subdisciplines of clinical and social psy­ chology, as well as the long-recognized role of social factors in the assessment, etiology, maintenance, and treatment of dysfunctional behaviors (Brehm, 1976; Hill & Weary, 1983; Leary & Miller, 1986; Weary & Mirels, 1982), it is not surprising that there is now a considerable literature at the interface of social cog­ nition and clinical psychology. Indeed, a letter we sent in preparation for this chapter to 85 scholars whose work lies at this interface yielded well over 500 relevant theoretical and empirical papers published during the last 5 years. One can only imagine (for us, with some horror) what a more systematic search would have yielded. In general, our search revealed evidence that social cognition has had at least some influence in most major areas of clinical psychology. For example, there is work on social cognitive processes involved in clinical assessment and deci­ sion making (e.g., Dawes, 1986; Fincham & Bradbury, 1991a; Jordan, Harvey, & Weary, 1988; Kayne & Alloy, 1988; Nezu & Nezu, 1989; Salovey & Turk, 1988, 1991; Turk & Salovey, 1986). There also is relevant work on psychother­ apeutic processes (e.g., Barlow, 1991; Brewin, 1989; Fincham & Bradbury, 1990a, 1991a, 1991b; Firth-Cozens & Brewin, 1988; Foersterling, 1985; Haaga & Davison, 1991; Kendall, Vitousek, & Kane, 1991; Rabinowitz, Zevon, & Karuza, 1989; Stoltenberg, Pace, & Maddux, 1986; but see Coyne, 1990). By far, the greatest amount of social cognition work in clinical psychology

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

291

is found in the general area of psychopathology and other forms of dysfunctional behavior and emotional distress (for recent, general treatments of this literature, see Horowitz, 1991; Maddux, 1991; Schwartz & Garamoni, 1989; Snyder & For­ syth, 1991; Taylor & Brown, 1988; Weary, Stanley, & Harvey, 1989; Wegner, 1992). There are now substantial literatures on the role of social cognition in the development and maintenance of, or adjustment to, marital conflict and vio­ lence (e.g., Andrews & Brewin, 1990; Fincham & Bradbury, 1990b, 1991a; Holtzworth-Monroe & Jacobson, 1985; Swann, Hixon, & De La Ronde, 1992), shyness and loneliness (e.g., Anderson & Arnoult, 1985; Anderson, Horowitz, & French, 1983; Anderson & Riger, 1991; Cutrona, 1982; Leary & Schlenker, 1981), maladaptive reactions to stressful life events (e.g., Brewin, 1984; Col­ lins, Taylor, & Skokan, 1990; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Janoff-Bulman, 1979; Janoff-Bulman & Lang-Gunn, 1988; Lefcourt, Martin, & Saleh, 1984; Linville, 1987; Perloff, 1983; Schulz & Decker, 1985; Taylor, Buunk, & Aspinwall, 1990; Taylor & Schneider, 1989; Thomp­ son & Janigian, 1988; Wood, Taylor, & Lichtman, 1985), and low self-esteem (e.g., Epstein, 1992; Pelham & Swann, 1989; Showers, 1992a, 1992b; Solo­ mon, Greenberg & Pyszczynski, 1991; Swann, Stein-Seroussi, & McNulty, 1992). Additionally, there are emerging literatures focusing on social cognitive fac­ tors involved in addictive behaviors (e.g., Banaji & Steele, 1989; Bradley, Gossop, Brewin, Phillips, & Green, 1992; Cooper, Russell, & George, 1988; DiClemente, 1986; Hull & Bond, 1986; Hull & Young, 1983; Lyon & Green­ berg, 1991; Stacy, Newcomb, & Bentler, 1991, 1993; Stacy, Widaman, & Marlatt, 1990; Vuchinich & Tucker, 1988) sexual abuse (e.g., Shapiro, 1989; Silver, Boon, & Stones, 1983), mental retardation (e.g., Gibbons, 1985), and schizophrenia (e.g., Brewin, MacCarthy, Duda, & Vaughn, 1991). However, it is probable that the largest literatures integrating social cognition and psychopathology are concerned with the emotional disorders: anxiety and depression. A computer search yielded about 245 relevant anxiety and 620 depres­ sion journal articles appearing since 1983. Such empirical attention should not be surprising; we have had prominent social-cognitive theories of depression for nearly two decades and of anxiety for nearly a decade. Moreover, depression and anxiety are two of the most prevalent mental health problems of today (Wil­ liams & Poling, 1989), and they frequently co-occur (Clark, 1989; Maser & Cloninger, 1990). Accordingly, in this first Handbook chapter on the interface of social cognition and clinical psychology, we focus on these two disorders. In particular, we examine how anxious and depressed perceivers process in­ formation in social situations, especially how they process information about other people. Such a focus seems important for at least three reasons. First, there is little doubt that individuals’ perceptions of others guide their social interactions in a significant way; because much has been written about the role of poor or inadequate social relationships in the development and maintenance of anxiety (Merluzzi, Rudy, & Krejci, 1986; Monroe & Wade, 1988) and depression (Blatt,

292

WEARY AND EDWARDS

Wein, Chevron, & Quinlan, 1979; Brown & Harris, 1978; Oatley & Bolton, 1985; Warren & McEachren, 1983; Weissman & Paykel, 1974), it seems essential that we have a clear picture of the impact of such conditions on social perception processes. Second, just as some of the best empirical evidence regarding the bas­ ic relationship between memory and the self came from studies of various pathol­ ogies (Greenwald & Pratkanis, 1984), examination of the way in which anxious and depressed individuals perceive, understand, and judge others may illuminate basic cognitive processes involved in person perception (cf., Averill, 1988). Fi­ nally, although there are a number of recent reviews of depressed and anxious individuals’ processing of personal information and the etiological significance of such processing (Alloy, 1988; Barlow, 1991; Eysenck & Mathews, 1987; Ken­ dall & Watson, 1989; MacLeod & Mathews, 1991; Mineka & Tomarken, 1989), we were unable to find a single, comprehensive review of the growing literature on other-perception processes associated with these disorders. In this chapter, we examine two general classes of factors that might play an important role in the social perception processes of anxious and depressed in­ dividuals, namely self-representations and motivational concerns. More specifi­ cally, we examine the possible role of depressed and anxious schemata and self-state representations in the perception and evaluation of others, paying par­ ticular attention to studies that permit a direct comparison of these populations. We also examine the role of control and social comparison motives in the per­ ception of others, because both motivations have figured prominently in a num­ ber of the contemporary theories of depression and anxiety. However, before turning to an examination of the empirical research, it may be helpful to review briefly three of the dominant cognitive theories that provide explanations for both depression and anxiety. Although these theories do not ex­ plicitly address to any great degree the role of depression or anxiety in social perception processes, they do provide insights into the nature of the self­ representations and motivational concerns characteristic of depressed and anx­ ious individuals. In so doing, they provide a critical starting point for our exami­ nation of depression, anxiety, and the processing of social information.

COGNITIVE THEORIES OF DEPRESSION AND ANXIETY 1 Beck's Schema Theory

The central construct in Beck’s theory is that of schema. A schema has been de­ fined as an enduring cognitive structure that represents an individual’s organized knowledge about a given concept or stimulus domain (Fiske & Taylor, 1984). !In this chapter, our focus is on trait, rather than state (induced or naturally occurring), concep­ tions of depression and anxiety. The reasons for this focus are several. First, although some researchers studying social cognition and depression have begun to make a distinction between state and trait

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

2 93

It contains a specification of the principle attributes of that domain, the relation­ ships among attributes, and specific examples or exemplars of it (Singer & Salovey, 1991). A schema guides attention to new information, the retrieval of stored in­ formation, the integration of information, and the generation of interpretations and inferences about (Brewer & Nakamura, 1984; Fiske & Taylor, 1984; Taylor & Crocker, 1981), as well as affective and behavioral responses toward, a stimulus. According to Beck’s schema theory of anxiety and depression (1976; Beck & Emery, 1985), the cause of the emotional disorders is a problem in the regulato­ ry functions of the cognitive system (i.e., idiosyncratic schemata are activated and remain hypervalent, whereas schemata inconsistent with these schemata re­ main deactivated). The idiosyncratic schemata of particular concern to Beck con­ tain negative information about the physical attributes, behaviors, and traits of the self and the self in relation to others (Cantor, Mischel, & Schwartz, 1982; Markus, 1977; Singer & Salovey, 1991). According to Beck, these negative sche­ mata concerning the self are developed early in life in response to certain situa­ tions and later are activated principally by negative life events presumably reminiscent of the earlier situations that gave rise to the schemata. Once activat­ ed, these schemata are thought to lead to systematic biases and distortions (e.g., minimization, selective abstraction, magnification, catastrophizing) in interpre­ tations of self-relevant information in a manner consistent with the schemata. Spe­ cifically, these biases and distortions are thought to lead to a negative cognitive set concerning the self, the future, and the world. The primary difference between anxiety and depressive disorders, Beck ar­ gued, resides in the specific content of the schemata. Depressogenic schemata contain prototypical information about personal loss, hopelessness, and failure, whereas the content of anxious individuals’ schemata revolves around themes of personal physical or social danger, vulnerability, and uncertainty. Although these

effects, for the most part there is no consensus with respect to which measures tap state as opposed to trait symptoms. Moreover, for both depression and anxiety, the major cognitive models predict and the available empirical evidence supports an interaction for state and trait variables, such that heightened mood state is associated with the greatest schema effects in individuals with the greatest corresponding disposition (MacLeod & Mathews, 1991); that is, although emotional states may acti­ vate relevant schemata, the degree to which such schemata are elaborated and produce the most marked effects depends on the level of trait emotion. We also do not focus on different diagnostic subclassifications of depressive and anxiety disor­ ders, because the cognitive biases generally are assumed to be operative across the different anxiety and depressive disorders (Abramson et al., 1989), but, at least in the case of anxiety disorders, manifest­ ed differently depending on the nature of the focal concern (e.g., fear of social situations vs. fear of snakes). Additionally, in much of the literature to be reviewed, the subject populations were subclinical populations evidencing mild to moderate levels of symptomatology (most often assessed with the BDI or some similar self-report inventory) for which no diagnosis was available. Where patient populations were employed or where the severity of symptomatology was relevant to the theoretical predictions, we have noted these points and have indicated the relevant levels or diagnosis.

294

WEARY AND EDWARDS

schemata may remain inactive for long periods of times, they become hypervalent during episodes of anxiety or depression and actively influence the process­ ing of social information. For example, depressogenic schemata should lead the depressed individual to attend to and recall information that is congruent with the schema, and to ignore, forget, or distort information that is incongruent with it. They also should result in depression-congruent interpretations of ambiguous situations. Similarly, the danger schemata of the anxious individual should result in greater attention to and recall of information relevant to physical or social threats and to resolution of ambiguity in favor of danger-related interpretations. The Helplessness-Hopelessness Model

Recently, Alloy and her colleagues extended and modified the hopelessness the­ ory of depression (Abramson, Metalsky, & Alloy, 1989) to account for both anxi­ ety and depression, as well as the comorbidity of these two disorders (Alloy, Kelly, Mineka, & Clements, 1990). Like Beck’s model, the helplessnesshopelessness model of anxiety and depression is a cognitive diathesis-stress model, where the cognitive diathesis entails negative beliefs about one’s outcomes. Brief­ ly, it posits the presence of negative life events and individual differences in at­ tributional and control styles as distal contributory causes of anxiety and depression. According to the model, the predisposition to explain important nega­ tive events in a way that implicates stable (i.e., enduring) and global (i.e., likely to affect many outcomes) factors increases the likelihood that individuals will develop a negative outcome expectation (a belief that negative outcomes are like­ ly to occur or that positive outcomes are not likely to occur). The predisposition to view negative events as uncontrollable increases the likelihood that individu­ als will develop a helplessness expectancy (a belief in the noncontingency of responses and outcomes). The model further posits that whether anxiety and/or depression results de­ pends on the interrelation of these expectations and the certainty with which they are held. Specifically, pure anxiety results from an uncertain expectation that one is helpless, whereas mixed anxiety-depression results from a certain expectation that one cannot control future outcomes combined with uncertainty about the likely occurrence of negative outcomes. Hopelessness depression results when this un­ certain negative outcome expectancy becomes certain (i.e., when one is certain that one cannot control future outcomes and that negative outcomes are certain to occur). Thus, the helplessness-hopelessness theory argues for a sequential relation­ ship between anxiety and depression. Such an argument is consistent with Wortman and Brehm’s (1975) integration of reactance and helplessness formulations, and with studies examining the temporal relationship between these disorders (Al­ loy et al., 1990). Finally, we should note that a substantial body of literature supports the notion that both anxious and depressed individuals display expecta­

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

295

tions of uncontrollability (for reviews see Alloy et al., 1990; Weary, Marsh, Gleicher, & Edwards, 1993). Self-Discrepancy Theory

Higgins (1987) recently presented a model that seeks to relate various kinds of self-belief incompatibilities with distinct kinds of emotional problems. In his self­ discrepancy theory, Higgins identified domains of the self and standpoints on the self as two cognitive dimensions underlying self-state representations. The first dimension, domains of the self, refers to the self-concepts related to (a) the “ac­ tual” self, or the person’s representation of the attributes that he or she actually possesses; (b) the “ideal” self, or the person’s representation of the attributes that he or she ideally would like to possess (aspirations, hopes, or goals); and (c) the “ought” self, or the person’s representation of the attributes that he or she should possess (rules, obligations, or prescriptions). The second dimension, stand­ points of the self, refers to the point of view from which a person is judged, namely the person’s own standpoint or the standpoint of significant others (e.g., mother, father, teacher, lover). In crossing these two dimensions, two combinations of domains of self and standpoints on self, actual/own and actual/other, comprise what typically is meant by self-concept. The other four combinations (ideal/own, ideal/other, ought/own, ought/other) constitute self-directive standards or guides. According to the the­ ory, individuals strive to achieve a match between their self-concept and one or more of their self-guides. When discrepancies arise, negative emotions result, and individuals are motivated to reduce the discrepancy. Moreover, the magni­ tude of the discrepancy, defined as the total number of attributes in the two self­ representations that are mismatched minus the number of matched attributes, de­ termines the intensity of the negative affect. Most important from our perspective is the additional notion that particular kinds of self-state discrepancies produce different kinds of negative affects. In particular, anxiety is thought to arise from discrepancies associated with actu­ al/own versus ought (own or other) conflicts, whereas depression should result from actual/own versus ideal (own or other) conflicts. That is, actual versus ought discrepancies psychologically represent the actual or anticipated presence of nega­ tive outcomes and the agitation-related emotions of fear, danger, vulnerability, uncertainty, and anxiety. Actual versus ideal conflicts denote the actual or ex­ pected absence of positive outcomes and are associated with the dejection-related emotions of dissatisfaction, disappointment, hopelessness, and depression. To the degree that these actual versus ought and actual versus ideal self-representations are activated frequently, they will be readily accessible and result in chronic dis­ comfort and emotional symptoms. Although the major focus of self-discrepancy theory is the prediction of specific kinds of emotional consequences of possessing specific kinds of self-belief in­

296

WEARY AND EDWARDS

compatibilities, it is important to note an additional consequence of possessing such cognitive constructs. Specifically, such available and chronically accessible cognitive representations are likely to influence the interpretation of and memory for ambiguous events, as well as negative events similar in content to the self­ discrepancy. Summary

All three of these theories of anxiety and depression argue for the existence of cognitive elements (self-schemata and self-state representations, attributional and control styles) that bias the processing of self-relevant personal and social infor­ mation and increase the likelihood that these emotional disorders will occur.2 Moreover, the proposed contents of these cognitive components bear some im­ portant thematic similarities to one another. That is, in depression-related struc­ tures, all of the theories implicate self-beliefs that one has lost or never will obtain desired outcomes, because of personal inadequacies and failures. In anxiety-related structures, all of the theories highlight self-beliefs that something terrible or dan­ gerous might happen. Before examining the role of such self-representations on processes of person perception, we review briefly some of the major findings from work seeking to examine the existence of such cognitive constructs.

SCHEMATIC BIASES IN DEPRESSION AND ANXIETY

Work that seeks to verify the existence of structural cognitive elements in anxie­ ty and, particularly, depression is voluminous. As outlined previously, these ele­ ments typically are postulated to be thoughts whose content is of a particular type: loss related in depression and danger related in anxiety. These cognitive elements, in turn, are thought to be related to specific processing biases. In this section, we examine the cognitive content that is thought to exist in these two disorders and review some of the processing biases thought to be linked to this content. The picture regarding processes specific to anxiety and depression is compli­ cated, because, as previously noted, depression and anxiety tend to co-occur. Thus, a process thought to be characteristic of one may be due to the fact that the other disorder also is present, such that either the other disorder is driving the effect or some type of interactive process is at work. Studies offering adequate direct comparison of processes unique to the two disorders have been rare (Alloy et al., 1990; Garber, Miller, & Abramson, 1980; Gotlib, 1984; Ingram, 1989; MacLeod & Mathews, 1991). 2Evidence regarding the important causal notions of these theories is inconsistent, and some researchers have argued that they have not been subject to adequate tests (e.g., for reviews see Al­ loy, Hartlage, & Abramson, 1988; Haaga, Dyck, & Ernst, 1991). However, a discussion of this evidence is beyond the scope of this chapter and does not bear on the major questions to be addressed.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

297

Thought Content

Depression. There has been substantial agreement on the nature of the thought content of depressives. As noted in our discussion of the major cognitive the­ ories, depression is characterized by thoughts that are deprecatory and pessimis­ tic with regard to the self, the future, and the world. A variety of methodologies have been used to verify the frequency, intensity, and valence of depressives’ thoughts: questionnaires (Beck, Brown, Steer, Eidelson, & Riskind, 1987; Clark, 1986; Crandell & Chambless, 1986; Hollon & Kendall, 1980), thought listings and diaries (Beck, Rush, Shaw, & Emery, 1985; Higgins, Bond, Klein, & Strau­ man, 1986), and endorsement of affectively toned adjectives regarding the self or the world (Derry & Kuiper, 1981; Dobson & Shaw, 1987; Greenberg & Beck, 1989; Roth & Rehm, 1980; Segal, Hood, Shaw, & Higgins, 1988; Strauman, 1990). It consistently has been found that depressives show increased negative thinking regarding the self (Alloy, 1988; Haaga, Dyck, & Ernst, 1991). In addition to these negative self-thoughts, depressed individuals also have been hypothesized to have certain types of dysfunctional attitudes. Specifically, depres­ sives have been found to have rigid and unrealistic beliefs about the rules for evaluating their self-worth (Beck et al., 1985; Dobson & Shaw, 1987; Kuiper & Olinger, 1986). Anxiety. The content of anxious individuals’ thoughts bears some similarity to that of depressives. As is true of depression, anxiety is associated with selfdeprecatory thoughts. Both groups report thoughts of helplessness and negative self-evaluation. However, anxiety is distinctively characterized by thoughts con­ cerning social or physical danger (Beck, 1967; Beck & Emery, 1985; Mathews, 1990). These thoughts sometimes center around the ramifications of the self as anxious (Holloway & McNally, 1987) or on questions about future outcomes (In­ gram & Kendall, 1987). In addition, anxious individuals have beliefs that nega­ tive events are especially likely (Craske & Barlow, 1991; Mathews, 1990; see also Garber et al., 1980). Like research on depression, research examining the thought content of anxious individuals has used a variety of methodologies, in­ cluding interviews (Beck, Laude, & Bohnert, 1974; Hibbert, 1984), question­ naires (Beck, Epstein, Brown, & Steer, 1988), self-endorsement of trait adjectives (Greenberg & Beck, 1989), and diaries (Sewitch & Kirtsch, 1984). Although studies describing the content of depressed and anxious individuals’ thoughts show results consistent with the major cognitive models of these disor­ ders, many cognitive theories of depression and anxiety go a step further and postu­ late an organized structural relationship between these cognitions. As noted earlier, Beck’s theory of depression explicitly postulates such a depressogenic schema, as does the model of Kuiper and colleagues (Kuiper, Derry, & MacDonald, 1982; Kuiper & Olinger, 1986; Kuiper, Olinger, & MacDonald, 1988). A similar cog­ nitive organization is apparent in the helplessness-hopelessness model. The anxiety

298

WEARY AND EDWARDS

models of Beck (Beck & Emery, 1985) and Ingram and Kendall (1987; Kendall & Ingram, 1987) also postulate “anxious schemata.” Postulating a schematic organization to the content of depressed and anxious individuals’ thoughts leads to a variety of hypotheses about the manner in which depressed and anxious persons deal with information. Research examining the cognitive processes particular to depression and anxiety can be broken roughly into three categories: biases in attention, biases in comprehension or interpreta­ tion, and memory biases. We only briefly review the major findings of this liter­ ature in this section. Although some of it is reviewed in more detail in a later section, the interested reader is referred to the various reviews cited here. Before turning to this literature, however, several general points should be made. First, the effects cited in the following literature pertain to subjects screened for depression (either at subclinical or clinical levels; see footnote 1), rather than to subjects whose mood has been manipulated. Second, the precise mechanisms underlying these effects often are unclear. That is, although the findings often are consistent with schema models of depression, they frequently are open to al­ ternative interpretations. This is partly due to the types of research methodolo­ gies that have been used (cf. Segal, 1988, and the following discussion). However, it also is due partly to underspecification of the schema construct in these models. In particular, detailed delineation of the interaction between cognitive content and processes involved in depression has been lacking (cf. Ingram, 1990a; Wil­ liams, Watts, MacLeod, & Mathews, 1988), at least as compared to the concep­ tualizations found in mainstream cognitive psychology (Segal, 1988; Strauman, 1992; Zuroff, 1992). Indeed, progress toward verification of the cognitive struc­ tures specific to depression and anxiety (e.g., schemata) is unlikely in the ab­ sence of strong models of the processes that accompany such content (cf. Anderson, 1978; Barsalou, 1990). Cognitive Processes in Depression

Attention. Research on attentional biases in depression are of two types. Con­ sistent with a schema model of depressive cognition, some theorists have postu­ lated that depressives engage in selective monitoring of negative (i.e., mood-congruent) information (Beck, 1967; Rehm, 1977). Studies that examine this notion typically employ designs testing attentional orientation toward nega­ tive words, or they use lexical decision paradigms. Studies of these types gener­ ally fail to find evidence of attentional biases for negative information (Ingram, 1990a; MacLeod & Mathews, 1991). There is some evidence that depressives will show disruption of performance when presented with negatively toned stimuli on modified Stroop (1938) tasks, although it may be that concomitant anxiety is driving the effects (Gotlib & McCabe, 1991; MacLeod & Mathews, 1991). The second type of attentional bias that has been examined in depressives is self-focused attention (Duval & Wicklund, 1972). Research consistently supports

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

299

the proposition that depression is associated with a tendency to direct attention toward oneself and away from the external environment (Ingram, 1990b; Pyszczynski, Hamilton, Greenberg, & Becker, 1991). Indeed, there currently are sever­ al models positing a central role for self-focused attention in depression (Carver & Scheier, 1981; Lewinsohn, Hoberman, Teri, & Hautzinger, 1985; Pyszczynski & Greenberg, 1987). These models propose that self-focused attention medi­ ates many of the features typically associated with depression, such as the tendency toward low self-esteem, increased self-evaluation, negative affect, withdrawal after failure, and internal attributions for failure (Smith & Greenberg, 1981). Although there is some evidence for the mediational role accorded self-focused attention in these models, more research clearly needs to be done (Musson & Alloy, 1988; Pyszczynski, Greenberg, Hamilton, & Nix, 1991; Pyszczynski et al., 1991). Self-focused attention appears to be found in several clinical disorders, and it is therefore not specific to depression (Ingram, 1990b). Thus, the important role of self-focused attention in depression is probably the way that it interacts with other elements of the depressive condition (e.g., self-directed attention may activate a depressogenic schema; Ingram, 1990a). Comprehension. Depressed individuals exhibit a negative bias in their in­ terpretation of ambiguous stimuli, as would be predicted by a schema model of depression. This effect has been found in questionnaires, judgments of individu­ als’ own performance, recall of feedback, and conclusions drawn from hypothet­ ical scenarios (Coyne & Gotlib, 1983; Haaga et al., 1991; MacLeod & Mathews, 1991). However, studies examining depressives’judgments of their own perform­ ance have been problematic. Although some researchers have concluded that depressives are actually more accurate in their self-appraisals than are non­ depressed individuals (the “depressive realism” effect; Alloy & Abramson, 1979, 1988; Ruehlman, West, & Pasahow, 1985), the findings in this area conflict, and a number of methodological concerns about the paradigm used to demon­ strate such realism effects have been raised (Ackermann & DeRubeis, 1991; Coyne & Gotlib, 1983; Segal & Shaw, 1986). For instance, there often is no clear or objective standard of accuracy in these studies, due to the existence of multiple standards that might be used by subjects (e.g., probability of success at specific experimental tasks vs. probability of success “in life”) or to inherently subjective dependent variables (e.g., confidence ratings). Memory. Schema models of depression make a variety of predictions regard­ ing depressed individuals’ memory processes. One consistently supported predic­ tion is that depressives show a memory advantage for negative (i.e., mood-congruent) information (Blaney, 1986; Coyne & Gotlib, 1983; Haaga et al., 1991; Ingram & Reed, 1986; MacLeod & Mathews, 1991). This effect oc­ curs when the material to be remembered entails personal memories and when

30 0

WEARY AND EDWARDS

it involves lists of differentially valenced words, although the latter class of ef­ fects seems to be most robust when the words are encoded with respect to the self (Dalgleish & Watts, 1990; Segal, 1988; Segal & Shaw, 1986). Evaluation o f Evidence for the Depressogenic Schema. Although much of the research examining depressives’ cognitive processes has been aimed at validat­ ing the depressogenic schema notion, and although much of the evidence is con­ sistent with the notion of such a schema, there is controversy regarding the strength of the current evidence and the adequacy of existing paradigms to test the struc­ tural aspects of depressive cognitions (Coyne & Gotlib, 1983; Segal, 1988). Most of the findings in this area can be explained readily by a construct accessibility account (Segal, 1988) or even a self-presentational (i.e., willingness to endorse certain types of items) interpretation (Coyne & Gotlib, 1983). As several the­ orists have noted (Coyne & Gotlib, 1983; Segal, 1988; Spielman & Bargh, 1990), the essential element of a schema that sets it apart from a network account of cognition is some type of organized structure. Research showing increased ac­ cessibility of negative content, particularly when those effects are limited to in­ dividuals currently experiencing negative moods, cannot distinguish effects due to an organized schema from mood congruent spreading activation along a sim­ ple associative network (Blaney, 1986; Segal, 1988). There have been some ef­ forts to use more appropriate paradigms to verify schematic processing (borrowed from cognitive psychology), such as recall of material encoded with reference to the self, reaction time, and decision consistency across self-judgments (Kuiper et al., 1988). However, interpretation of these results remains problematic (Bargh & Tota, 1988; Segal, 1988; Spielman & Bargh, 1990). Additionally, lon­ gitudinal studies of depressive self-schema have yielded results more consistent with mood congruency than a self-schema account (Hammen, Marks, deMayo, & Mayol, 1985). Research examining automatic processing in depression regard­ ing the self (e.g., unintentional diversion of attention toward schema-congruent material in the Stroop task, semantic priming paradigms) offers promising alter­ natives. However, there are few of these studies, and the results so far have been interpreted as more consistent with a construct accessibility interpretation (Moretti & Shaw, 1989; Spielman & Bargh, 1990). For the current discussion, the issue of whether self-elements are organized and functionally linked in memory is rela­ tively unimportant; both construct accessibility and schema models argue that self-constructs, once activated, exert an influence on people’s processing of so­ cial information. Finally, in defense of the schema notion, depression theorists have proposed a variety of modifications of schema theory: abandoning the extreme structural assumptions (Ingram, 1990a); adding the notion that the depressogenic schema is dormant when the individual is not in a depressive episode (Beck, 1967; In­ gram, 1990a; Segal & Shaw, 1986); and more generally moving toward an inter­ active account based on multiple cognitive constructs interacting with environ­

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

301

mental stressors (Beck, 1967; Haaga et al., 1991; Metalsky, Abramson, Seligman, Semmel, & Peterson, 1982; Segal & Shaw, 1986). Cognitive Processes in Anxiety

Attention. There is strong evidence for an attentional bias in anxiety. Research using the Stroop task found that threatening stimuli that match the specific wor­ ries of anxious subjects interfere with subjects’ task performance. More direct measures of attentional orientation have shown that anxious subjects appear to allocate attention automatically to anxiety-specific threat cues in the environment (Dalgleish & Watts, 1990; Litz & Keane, 1989; MacLeod & Mathews, 1991; Mathews, 1990). These tasks typically require subjects to attend to some stimuli (e.g., dot probes appearing at various areas on a video screen) while distractor material (e.g., words) that is either neutral or threat-relevant is presented simul­ taneously. Attention to the distractor is measured via response latencies to detect the probe. Evidence of self-focused attention also has been found in anxiety, possibly as a result of the increased arousal that accompanies the identification of threat. However, despite a large amount of theoretical support, relatively little research has been conducted on the role of self-focused attention in anxiety (Craske & Barlow, 1991; Hope, Gansler, & Heimberg, 1989; Ingram, 1990b). Comprehension. Research on comprehension processes in anxiety has cen­ tered around the notions that anxious individuals overestimate the probability of negative events and tend to interpret ambiguous stimuli as threatening. The ex­ istence of both of these biases have been supported (MacLeod & Mathews, 1991; Mathews, 1990; Mineka & Tomarken, 1989). For instance, anxious compared with control subjects have been shown to rank threatening interpretations of am­ biguous scenarios as more likely (Butler & Mathews, 1983) and to spell ambigu­ ous homophones (e.g., grown/groan) in a threat-consistent manner (Eysenck, MacLeod, & Mathews, 1987). Memory. Although models of anxiety typically postulate that anxious people should show threat-congruent recall biases, there is very little evidence consis­ tent with this proposition. In fact, there is even evidence to suggest that anxious individuals have poorer memory for anxiety-relevant material (Craske & Barlow, 1991; Litz & Keane, 1989; MacLeod & Mathews, 1991; Mathews, 1990; Mogg, Mathews, & Weinman, 1987; Pyszczynski et al., 1991). These findings are some­ what difficult to reconcile with the findings regarding attention, because it would be expected that greater attention would have some enhancing effect on memory. One explanation for this is that attentional allocation in anxiety is an automatic process, but that once anxious individuals have identified and attended to a threat, they tend to avoid further conscious elaboration of the information, which tends

302

WEARY AND EDWARDS

to suppress recall (Mathews, Mogg, May, & Eysenck, 1989; Williams et al., 1988). Summary

A great deal of descriptive research has been conducted on depressogenic and anxiety schemata. By and large, the findings are consistent with the existence of such schemata, although numerous important questions remain. Throughout the remainder of this chapter, we examine several cognitive and motivational fac­ tors that may influence depressed and anxious individuals’ processing of infor­ mation about other people, rather than processing of information about the self. Let us turn first to a consideration of important cognitive influences.

SCHEMATIC EFFECTS IN THE PERCEPTION OF OTHERS

A myriad of social-cognitive constructs and categories have been shown to in­ fluence social perception processes (Brewer, 1988; Fiske & Neuberg, 1990) of normal individuals. For example, stereotypes, perceivers’ implicit personality the­ ories, role schemata, person schemata, and attributional and other procedural sche­ mata (Fiske & Taylor, 1984; Kelley, 1972; Reeder, 1985) all, when they are active, may influence social perceivers’ understanding and impressions of others. A particularly important cognitive element for the purposes of the present dis­ cussion is the self (Holyoak & Gordon, 1983; Srull & Gaelick, 1984). Theorists long have suggested that an implicit, if not explicit, comparison of self with others is an integral part of self-perception, as well as other-perception (Festinger, 1954; Heider, 1958; Snygg & Combs, 1959). Several lines of research typically are offered as demonstrations that the self may influence the perception, construal, evaluation of, and memory for another person. For example, research on the self­ based consensus, or false consensus, bias has indicated that perceivers use their own attitudes, characteristics, and preferences as reference points for estimating the frequency and judging the appropriateness of another’s attitudes, characteris­ tics, and preferences (Ross, 1977). Additionally, research has shown that per­ ceivers often describe others in terms of important or central self-relevant traits (Fong & Markus, 1982; Higgins, King, & Mavin, 1982; Lemon & Warren, 1974; Lewicki, 1983; Markus & Smith, 1981; Shrauger & Patterson, 1974); they also attend to and seek more information about others that is relevant to their own self-schemas (Fong & Markus, 1982) and sometimes demonstrate better immediate recall of another’s specific, schema-relevant behaviors (Catrambone & Markus, 1987). However, other research has not found consistent, pervasive, or large effects of self-structure on other-perception (Hamilton, Katz, & Leirer, 1980; Park, 1986; Prentice, 1990). These findings have resulted in several attempts to specify the

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

303

conditions under which the effects of the self-structure are most pronounced. For example, one proposal that has received some empirical support is that the perceiver’s self-representation is more likely to influence the perception of others when the perceiver is required by virtue of (a) the judgment task, (b) the perceiver’s expectations, (c) the relationship between the perceiver and the target (e.g., an outcome-dependent relationship or close friendship), or (d) impoverished target information to go beyond the data given. Thus, neither on-line, specific, evaluations of another, even on self-relevant traits (provided the target informa­ tion is readily available), nor immediate recall of another’s specific behaviors would be expected to reveal an influence of the perceivers’ self-schema. These tasks likely would focus the perceiver’s attention on the specific stimulus person. However, tasks requiring a general impression of the target, or a prediction of the target’s behavior in varied situations, should show robust effects of perceivers’ self-representations, particularly if the judgment is of some importance to the perceivers (Catrambone & Markus, 1987). In the preceding section, we saw that the self-representations of depressed and anxious individuals are charged affectively and contain more negative content than those of nondepressed, nonanxious individuals. Moreover, the empirical evi­ dence indicates that such self-representations exert an influence on the process­ ing of self-relevant, schema-congruent material, although the locus or stage of their effects may differ somewhat for the two disorders. In this section, we examine whether such self-representations also guide the processing of information about others. Specifically, do depressogenic and anx­ ious self-representations exert an influence on the selection and interpretation of and the memory for information about others? Are the affectively ladened self­ schemata of depressed and anxious individuals more accessible and, consequent­ ly, more influential in the perceptions of others, compared with those of nonanx­ ious and nondepressed perceivers? Under what conditions do anxious and depressogenic self-schemata influence other-perception? Aside from the possi­ ble influence of self-schemata in social perception, there is the question of the possibly different person schemata typically used by depressed and anxious per­ ceivers in construing others, compared with those of nondepressed, nonanxious perceivers and compared with each other. Although research on the social perception processes of depressed and anx­ ious populations is in its infancy, we review the available empirical evidence with an eye toward these questions. Just as there is more work on depressogenic than on anxious schemata, undoubtedly due to the relative recency of cognitive the­ ories of anxiety disorders, there also is more evidence relevant to depressives’ than anxious individuals’ social perception processes. We first review the studies employing depressives and then turn to a discussion of the research with anxious populations. Most of these latter studies also examined depressed groups of sub­ jects, and so permit a direct comparison of the social perception processes of these two groups.

304

WEARY AND EDWARDS

Depression and the Perception of Others

General Beliefs

Are depressives’ general beliefs about typical, hypothetical others more nega­ tive than nondepressives’ beliefs? Examination of the empirical evidence seems to indicate that they are not consistently or overwhelmingly more negative. For example, Swallow and Kuiper (1987) found that there were no differences in the content of free descriptions of the attributes of a typical other produced by groups of subjects presumably at risk for the development of depression (individuals with a large number of dysfunctional attitudes) and normal control subjects. Instead of free descriptions, Shapiro (1988a) had depressed and nondepressed college students evaluate people in general (Americans approximately the same age as the subjects) on 36 person-descriptive adjectives. The evaluative domains covered by these adjectives included morality and kindness toward others, achievementrelated abilities, social skills, and subjective well-being. Shapiro reported that, for men, there was a relatively weak, albeit significant, relationship between over­ all level of depression and negative perceptions of others; but, for women, there was no relationship between these variables. Only when specific dimensions (de­ pendency, self-criticism, and efficacy) of depression (Blatt, Quinlan, Chevron, MacDonald, & Zuroff, 1982) were examined was there some limited evidence that subjects’ self-representations generalized to the perceptions of others. Selfcritical men and dependent women evaluated others more negatively overall, although not necessarily just on evaluative dimensions corresponding to the con­ tent of their self-representations. However, in another study, Shapiro (1988b) reported that, for men and women, there was no relationship between overall depression and general beliefs about the success of people in general. When the specific dimensions of depression were examined, the only consistent effects were for dependent women, who saw others as failing to achieve desired outcomes more often. Additionally, several studies (Bargh & Tota, 1988; Kuiper & MacDonald, 1982) examined the content of depressed and nondepressed subjects’ self- and other-person schemata by having subjects judge whether a list of depression­ relevant and nondepression-relevant adjectives described themselves or others. Generally, in the self-reference conditions, both mildly and moderately depressed and nondepressed groups endorsed nondepressed-content adjectives to a greater degree than depressed-content adjectives. However, depressed subjects endorsed more depressed-content adjectives than nondepressed subjects. (It should be not­ ed that the balance or mix of positive to negative self-schema content may vary with the level of depression [i.e., clinically depressed subjects have been found to possess more negative content self-schemata, whereas the self-schemata of mild­ ly depressed subjects demonstrate a greater balance of positive and negative con­ tent; Derry & Kuiper, 1981; Kuiper et al., 1988]). For generalized others, the

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

305

content of the respective schemata appeared to be positive and equally so for both depressed and nondepressed subjects.3 In addition to judging the self- and other-descriptiveness of depressed-content and nondepressed-content adjectives, Bargh and Tota (1988) also assessed sub­ jects’ chronically accessible social constructs using a free-response measure de­ veloped by Higgins et al. (1982). This measure asks subjects to list up to 10 characteristics that they feel describe each of five different types of people: fre­ quently encountered, sought out, avoided, liked, and disliked others. Analyses of subjects’ responses to this measure indicated very similar content, both in terms of valence and depressed/nondepressed content, across all person types for the depressed and nondepressed groups. Far from indicating a general tendency for depressives to see others more nega­ tively than nondepressives, the previously cited research seems to suggest few consistent differences as a function of depression. There is some weak evidence from Shapiro’s studies that particularly central or important dimensions of depres­ sion may be related to specific negative beliefs about others. However, because of the inconsistencies in the results and numerous unexpected gender differences, firm conclusions about the generalization of specific self-beliefs to beliefs about others must await further research. Inferential Processes

The largest group of studies on depression and other-perception has focused on the effects of self-structures on various inferential processes.4 For example, studies have focused on judgments of social consensus for one’s attributes, future event predictions, social performance evaluations, attributions, judgments of con­ tingency, and decision styles and perceived risks for self and others. We discuss each of these literatures in turn. Attribute Consensus. Studies examining self-based consensus effects gener­ ally have found that individuals who hold a particular position or who have a particular attribute are more likely to view generalized or unknown others as also holding that position or possessing that attribute, compared with those who do 3For similar findings, see the evaluation results for depressed and nondepressed-control condi­ tion subjects in studies by Weary, Jordan, and Hill (1985) and Weary, Elbin, and Hill (1987). 4Our categorization of the preceding research as depicting the content of general beliefs about others and the following research as more relevant to inferential processes is, of course, somewhat arbitrary; the cognitive tasks in both sets probably involve some amount of inference. Still, the dis­ tinction was made for two reasons. First, the experimental tasks employed and the cognitive products examined in the following research seemed to require a more explicit focus on the cognitive integra­ tion of stimulus information or on cognitive elaboration. Second, we tried to be consistent with other reviews on depression and anxiety that have distinguished between the content of stored beliefs and various cognitive operations or processes (e.g., inferential or retrieval processes), and their categori­ zation of studies as relevant to one or the other of these.

306

W EARY AND EDWARDS

not (Ross, Greene, & House, 1977). Do depressed individuals, who presumably hold accessible, negative self-schemata, exhibit this same tendency to use their own attitudes and attributes as reference points for evaluating the frequency with which others possess the same attitudes and attributes? There have been at least five studies that have examined this question (Campbell, 1986; Crocker, Kayne, & Alloy, 1985; Fumham & Brewin, 1988; Kuiper & MacDonald, 1983; Tabachnik, Crocker, & Alloy, 1983), and all have yielded similar effects. In one of these, Tabachnik et al. (1983; see also a reanalysis by Crocker et al., 1985) asked depressed and nondepressed college students to rate the extent to which depression-relevant, nondepression-relevant, or depression-irrelevant attributes were true of themselves and were true of the average college student. They then examined three different aspects of depressed and nondepressed sub­ jects’ self-other comparisons: (a) the absolute differences between self- and otherratings; (b) the relative difference, within level of depression, in ratings of others between those who see attributes as relatively true of self and those who do not (the usual operationalization of the false consensus effect); and (c) the signed self-other difference ratings that reveal the extent of self-enhancement relative to others. The results of analyses on the absolute comparison index indicated that depressed compared with nondepressed subjects saw themselves as less similar to others. Moreover, the analysis of the relative difference index indicated a larger false consensus effect for nondepressed than for depressed subjects, particularly on depression-relevant attributes. In fact, depressed subjects showed no relation­ ship between self- and other-ratings on depression-relevant and nondepression­ relevant items when the items were combined into their respective scales.5 Fi­ nally, examination of the signed self-other differences indicated that nondepressed subjects consistently enhanced themselves relative to others, whereas depressed subjects did not (i.e., they sometimes enhanced and sometimes depreciated them­ selves relative to others). Although the Tabachnik et al. (1983) study did not permit an assessment of the accuracy of depressed and nondepressed subjects’ estimates of consensus, it was possible to do so in another study (Campbell, 1986), which focused on opin­ ion consensus. Campbell reported that, again, depressed college students were found to exhibit a lesser false consensus bias than nondepressed students. Among depressed students, this bias decreased with increasing levels of depressive symp­ tomatology. Importantly, analyses of the absolute accuracy scores revealed no effects associated with level of depression. On average, depressed subjects were not more accurate in their estimates than nondepressed subjects; they simply were less biased in the direction of their inaccuracies. (We return to the issue of depres­ sive accuracy in later sections of this chapter, and find that, on some judgments 5Although not a major focus of the study, results reported by Gibbons (1986) suggest that false consensus effects, at least for moods, might be increased when depressed individuals are in a state of heightened self-awareness.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

307

and under some conditions, depressed subjects appear more accurate in their self-, but not other judgments.) Future Event Predictions. Although the discussed studies suggested that there is no relationship between depressives’ self-representations and their perceptions of social consensus for their own attributes, the literature concerned with per­ ceptions of consensus for events tell a different story. There have been several studies (Alloy & Ahrens, 1987; Andersen, Spielman, & Bargh, 1992; Brewin & Fumham, 1986; Crocker, Alloy, & Kayne, 1988; Pyszczynski, Holt, & Green­ berg, 1987) that have asked depressed and nondepressed college students to esti­ mate the extent to which positive and negative events are likely to happen to themselves and to the average college student. The major cognitive theories of depression would predict that depressives’ views of the future should reflect pessimism relative to those of nondepressed individuals. They also would seem to predict that these pessimistic forecasts of the future should be limited to depressives’ predictions for themselves. For example, Beck (1967) argued that depressed people’s inferences reflect a “systematic bias against the self” (p. 234), and Abramson, Seligman, and Teasdale (1978) suggested that depressives are characterized more by personal than universal helplessness (see also Garber & Hollon, 1980). However, an examination of the empirical evidence reveals that depressed sub­ jects’ relatively more pessimistic predictions of event occurrences, particularly negative events, characterize both their predictions for self and for others. Although depressed and nondepressed subjects generally rate positive events as more likely to occur for themselves than negative events, this pattern is signifi­ cantly more pronounced for nondepressed subjects; depressed subjects show great­ er even-handedness in their self-estimates with increasing levels of symptoma­ tology. In terms of depressed-nondepressed comparisons, depressed subjects generally rate negative events as more likely to occur for themselves. For gener­ alized others, depressed compared with nondepressed subjects also generally view negative events as more likely to occur. These results seem to suggest that both depressives and nondepressives use their self-based knowledge in making frequency estimates for future event oc­ currence for both self and others. However, a recent study by Andersen et al. (1992) indicated that such implicit consensus information might be contained in a “future-event schema,” rather than a self-schema. The authors noted that there is an increasing tendency among mildly, moderately, and severely depressed in­ dividuals to think about the future and to do so in increasingly pessimistic ways. “Through experience and rehearsal, [moderately and severely depressed individu­ als] may acquire a highly efficient knowledge structure or schema for predicting future events, one that operates relatively effortlessly” (p. 5). To the degree that this processing efficiency exists in predicting future events, moder­ ately and severely depressed individuals’ future-event judgments should be made

308

WEARY AND EDWARDS

relatively automatically (i.e., with minimal cognitive effort) for positive and nega­ tive outcomes and for self and others. On the other hand, if such judgments are mediated by self-representations, Andersen et al. argued that we should then see moderately and severely depressed individuals making automatic judgments only when they concern negative events considered in relation to the self. Nondepressed individuals, of course, similarly would be expected to show automaticity only for their own positive future event predictions. The rationale for these predictions is that when the individual ac­ cesses the self-schema to make some particular judgment, the judgment should be made relatively automatically only if there is a match between the schema con­ tent and the to-be-judged stimulus. To test whether future event predictions are mediated by future-event sche­ mata or by self-schemata, Andersen et al. used a dual-task, concurrent memory load paradigm (e.g., Logan, 1988). This paradigm is based on a limited atten­ tional capacity notion. If a task is capacity-demanding, it should interfere with the simultaneous performance of another capacity-demanding task. Alternative­ ly, “. . . the less effort or attention required by a task (i.e., the more automated the task), the less it should be affected by constraints on the availability of atten­ tional resources” (Andersen et al., 1992, p. 713). Consistent with the results of other studies of depressive event prediction, An­ dersen et al. found that moderately depressed college students predicted more negative events and fewer positive events for both self and others than did mildly dysphoric (who presumably have not yet developed through experience and re­ hearsal a future-event schema) or nondepressed individuals. Moreover, consis­ tent with the future-event schema hypothesis, moderate depressives’ future-event predictions showed automaticity for both negative and positive events; that is, the introduction of a capacity-demanding task did not increase their response laten­ cies for future event predictions, but it did so for both mildly dysphoric and non­ depressed individuals.6 Although the bulk of the research on depressives’ future event predictions sug­ gests that depressives’ pessimism extends to the prediction of events for general­ ized others, and although one study suggests that these predictions may result from a future-event schema, there is one notable exception to the finding of general negativity in depressives’ other-event predictions. That exception is a study report­ ed by Pietromonaco and Markus (1985). In this study, depressed compared with nondepressed college students were better able to imagine negative events occur­ ring to themselves and estimated the likelihood of these events occurring in their lives as greater. However, this negativity in thought on the part of the depressed students did not extend to another individual (i.e., depressed subjects’ ratings of ^ h e s e results seem to argue against the Haaga et al. (1991) suggestion that Beck’s cognitive triad (negative beliefs about the self, the future, and the world) be revised and reduced to a single dimension, namely, negative view of self.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

309

their mental pictures and frequency estimates of negative events for another in­ dividual did not differ from nondepressed subjects’ ratings). There are two differences between this study and the other studies of depres­ sive event prediction that might help account for the inconsistency in results. First, although the studies showing a generalized negativity on the part of depressives employed within-subjects designs, the Pietromonaco and Markus (1985) study used a between-subjects design. However, this does not seem to be a particularly likely explanation, because between-subjects designs generally are thought to present subjects with a more ambiguous judgment task. The role of self­ representations or future-event schemata should be more pronounced in ambigu­ ous (between-subjects designs) than unambiguous conditions. The second differ­ ence was that the target other in the Pietromonaco study was a friend whom subjects had known for no more than 4 months; the target others in the studies showing generalized negativity were hypothetical, typical college students. It seems reasonable to expect that where there is a stored representation of the target other that contains information relevant to the judgment task, that representation, rather than a generalized future-event schema or even a self-schema, will be employed. Interpretations o f Social Relations. Several studies (e.g., Campbell & Fehr, 1990; Gotlib & Meltzer, 1987; Strack & Coyne, 1983) examined whether depres­ sives’ (or low self-esteem individuals’) interpretations of their social interactions are more negative than those of nondepressed individuals and, if so, whether these interpretations reflect self-schematic biases or are veridical perceptions. Of those studies that have obtained ratings of depressed and nondepressed subjects’ evalu­ ations of their own and their partners’ behaviors, there is consistent evidence that depressed college students evaluate their own social behaviors more negatively than do nondepressed students. However, these relatively negative self-evaluations on the part of depressives typically are not corroborated by interaction partners’ or by outside observers’ ratings. Whether subjects’ self-ratings are biased by highly accessible self-representations or are veridical perceptions is not an easy ques­ tion to answer, because it depends, among other things, on the criterion of ac­ curacy chosen (i.e., partners vs. observers who, for reasons that are not entirely understood, tend to be excessively harsh in their evaluations of both target sub­ jects and partners) and on how accuracy is defined (e.g., stereotypical vs. differen­ tial; see Campbell & Fehr, 1990; Kenny & Albright, 1987; Kruglanski, 1989; Swann, 1984, for discussions of the accuracy issue). Perhaps the most that can be said is that sometimes depressed individuals’ negative views of their social behaviors will be accurate, and sometimes they will not. Do depressed students’ relatively negative self-evaluations extend to their part­ ner ratings? The evidence seems to indicate that they generally do not; depressed and nondepressed students’ evaluations of their partners tend to be equally favora­ ble. There is, however, one report of results that were inconsistent with this general finding. Gotlib and Meltzer (1987) found that depressed compared with non­

310

WEARY AND EDWARDS

depressed students evaluated their interaction partners more negatively. However, the partners in this study also rated themselves more negatively. Consequently, the depressed subjects’ more negative evaluations of their partners in this study might have been based on actual behavioral differences during the interaction. There are two other studies (Hoehn-Hyde, Schlottmann, & Rush, 1982; Moretti & Miller, 1988) of depressed-nondepressed differences in interpretations of so­ cial behaviors that are relevant to the present discussion. In both of these studies, depressed individuals (college students and psychiatric patients) were asked to imagine positive and negative social behaviors being directed toward them or toward another person. The major results of these studies by now should not be surprising to the reader. In the self-reference condition, depressed compared with nondepressed subjects were either evenhanded and automatic in their processing of positive and negative behaviors (Moretti & Miller) or were more negative (Hoehn-Hyde et al.) in their reactions to negative behaviors directed toward them. In the other-referent conditions, depressed compared with nondepressed subjects demonstrated either a slight positivity bias in their processing of positive and nega­ tive information (Moretti & Miller) or showed no differences (Hoehn-Hyde et al.) in their reactions to positive and negative behaviors directed toward others. Wenzlaff and Berman (1985) reported similar findings (i.e., no depressednondepressed differences in reactions to positive feedback for both self or others, nor for unfavorable feedback given to others) to those of Hoehn-Hyde et al., although, instead of looking at social behaviors directed toward the self or others, they examined depressed and nondepressed students’ perceptions of favorable and unfavorable personality feedback. Attributions, Decision Styles, and Control Judgments. Because few studies have examined depressed-nondepressed differences for self and others in attri­ butions for positive and negative outcomes, in the assessment of benefits and risks in decisions, and in judgments of control or contingency over positive and nega­ tive outcomes, we summarize their major findings as a group. However, it is important to emphasize that firm conclusions from this group of studies must await replication of their results. To our knowledge, there is only one study (Sweeney, Shaeffer, & Golin, 1982) of depressed-nondepressed attributional style differences for self and for (gener­ alized) others and one study (Pietromonaco & Rook, 1987) of depressed-non­ depressed decisional styles for self and for (known and actual, but not well-known) others. In both studies, the subjects were college students. Moreover, the inves­ tigators of both studies expected that self-reference would be a critical compo­ nent of depressed students’ negativity. The results confirmed these predictions. Sweeney et al. reported that depressed compared with nondepressed students made higher composite internal, stable, and global self-attributions for negative out­ comes and lower composite attributions for positive outcomes; the two groups did not differ in their attributions for another’s negative or positive outcome.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

311

Similarly, Pietromonaco and Rook found that, within self-reference conditions, perceptions of risk in common life situations were correlated more strongly with decisions about the likelihood of engaging in target actions for depressed than for nondepressed students. Within the other-referent conditions, decisions made by depressed and nondepressed students were not correlated differentially with risk perceptions.7 By now, there is a reasonably large literature concerned with depressive real­ ism and nondepressive illusions of control in judgments of contingency. Although a number of important boundary conditions have been identified (Ackermann & DeRubeis, 1991; Alloy & Abramson, 1988), in general, depressed subjects ac­ curately judge their lack of control over positive and negative outcomes on ob­ jectively noncontingent problems, whereas nondepressed subjects generally overestimate their degree of control for positive (or high frequency) but not nega­ tive outcomes (but see Wyer and Srull, 1989, for an alternative interpretation of these results). Several studies (Martin, Abramson, & Alloy, 1984; Vazquez, 1987, Experi­ ments 3 and 4) examined whether the self-reference effects associated with level of depression extend to depressed and nondepressed individuals’ judgments of others’ control. “Are depressed individuals as accurate in judging how much control other people have over outcomes as they are in judging their own control? Do nondepressives succumb to an illusion of control when judging other people’s control over outcomes as well as their own?” (Martin et al., 1984, p. 126). As was the case for the attributional and decisional style studies reviewed earli­ er, the answer seems to be no. In judging another’s control, depressed subjects either appear to overestimate the degree of the other’s control for a noncontin­ gent positive outcome (Martin et al.), whereas nondepressed subjects’judgments are accurate, or the two groups of subjects are relatively and equally accurate (Vazquez, Experiment 4) in judging another’s noncontingent positive and nega­ tive outcomes. Memory Processes

Just as self- and other-schemata may exert an influence on the interpretation of information about self and others, so too may they influence memory for that information. Using an incidental free recall paradigm, several investigators (Bargh & Tota, 1988; Kuiper & MacDonald, 1982) examined whether depressed and nondepressed individuals recall more depressed-content compared with non­ 7Pietromonaco and Rook (1987) found depressed compared with nondepressed subjects to em­ phasize risks in their perceptions of life problems, whether they are thinking about themselves or about another person. These findings are consistent with the evidence cited earlier regarding the ex­ istence of future-event schemata (Andersen et al., 1992) and do not necessarily indicate any role of self-schemata in such judgments, particularly in light of the self-reference effects obtained on indices of the influence of risk perception on decisions styles.

312

W EARY AND EDWARDS

depressed-content adjectives previously rated as descriptive of self and others. The results of these studies indicate that, for self-judgments, clinically depressed subjects recalled more depressed- than nondepressed-content adjectives; mildly depressed subjects recalled equal numbers of, or more nondepressed- than depressed-content adjectives; and nondepressed subjects recalled more nondepressed- than depressed-content adjectives. Finally, for self-judgments, depressed compared with nondepressed subjects recalled more depressed-content adjectives. For judgments of others, depressed (mild and moderately) and non­ depressed subjects’ recall for nondepressed- and depressed-content adjectives were quite similar and generally favorable (although in the Kuiper & MacDonald study, nondepressed subjects recalled more depression-content adjectives for others than did depressed subjects). There is at least one other measure of schematic processing reported in the Bargh and Tota (1988) article that is highly relevant to the present discussion of depressed and nondepressed self- and other-schemata, although it is not a memory measure. This measure is an index of the decision latency for subjects’ responses to the depressed- and nondepressed-content adjectives. Although in­ terpretation of decision latencies is often complex and difficult, in this study the authors employed, in some conditions, a concurrent memory load task in an at­ tempt to examine the possible automaticity of self- and other-schemata. The authors reasoned that “. . . the load manipulation should have increased response laten­ cies less, relative to the no-load condition, to the extent that the social construct corresponding to the target item [adjective] became active automatically in the course of self- or other-reference” (p. 929). Consistent with this reasoning, the analyses of the latency measure indicated that, under self-referent conditions, the effects of memory load were smaller for depressed subjects judging depressedthan for nondepressed-content adjectives; the reverse was true for nondepressed subjects. For other-reference conditions, the effects of memory load were smaller for nondepressed- than depressed-content adjectives for both depressed and non­ depressed subjects. Thus, these results, along with the recall data, suggest that although depressed and nondepressed subjects have some self-schema content similarities, as well as some differences, they have very different self-schema content accessibilities. The latency and recall results also suggest that depressed and nondepressed subjects do not differ in the content or accessibility of their other-representations. Anxiety and the Perception of Others

General Beliefs

Two studies (Greenberg & Alloy, 1989; Vazquez & Alloy, 1987) examined the content of anxious college students’ beliefs about themselves and others. Both of these studies also included groups of depressed students. In these studies, groups

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

313

of subjects were identified by their responses to the Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) and the Trait Anxiety Inventory (TAI; Spielberger, Gorsuch, & Luschene, 1970). Students whose scores fell in the upper quartiles of the population on the BDI ( > 8, regardless of TAI score) and the TAI (> 37; BDI < 9) were placed in the depressed and anxious groups, respectively; students whose scores were below the cutoff points for both the BDI and the TAI were classified as normal, nondepressed-nonanxious. Con­ sistent with the theoretical formulations and evidence of the comorbidity and pos­ sible temporal sequence of anxiety and depression cited earlier in this chapter, the depressed group of students in these studies are best described as a mixed anxiety-depressed group; that is, the depressed group scored as high as the anx­ ious group on the TAI, but higher than the anxious group on the BDI. The content of these depressed, anxious, and normal students’ self- and otherschemata were examined by asking subjects to judge whether a list of positive and negative depression-relevant (DR), anxiety-relevant (AR), and nondepressednonanxious control adjectives described themselves and their best friends. Addi­ tionally, subjects in the Vazquez and Alloy (1987) study were asked to judge whether the lists of adjectives were true of people in general. This referent con­ dition was included because previous research found that self-other differences are greater when the self is compared with an unknown other about whom there is less extensive knowledge (Bower & Gilligan, 1979; Keenan & Baillet, 1980; Prentice, 1990). The general findings for the self-reference condition were as expected. There were no group differences in the endorsement rates for the control adjectives; all groups endorsed a high number of positive and a low number of negative con­ trol adjectives. For the positive DR and AR adjectives, depressed and anxious groups endorsed equal, but fewer, numbers of adjectives as true of themselves than did the normal group. For negative DR items, the normal group endorsed fewer adjectives than the anxious group, which, in turn, endorsed fewer adjec­ tives than the depressed group. For the negative AR items, the normal group endorsed fewer items than either the depressed or anxious group; the latter two groups did not differ in their endorsement rates of the negative AR items. Thus, for the self-reference conditions, both of these studies suggest that the normal group was characterized by highly positive self-schemata, whereas the anxious and depressed groups were characterized by relatively more negative self­ schemata, with the anxious groups evidencing some content specificity in their self-schemata. For both studies, the results for the best-friend-referent conditions revealed very few differences associated with groups. In both studies, students’ views of best friends were highly positive; all subjects consistently endorsed more posi­ tive than negative adjectives of all three types. In the Vazquez and Alloy (1987) study, subjects’ adjective endorsements for generalized others, although still posi­ tive, were not quite as positive as their best-friend schemata. Clearly, then, this

314

WEARY AND EDWARDS

research suggests that the negative contents of depressed and anxious individu­ als’ self-representations do not generalize to their conceptualizations of others. Inferential Processes

Attribute Consensus. We only could find one study that examined self-based, or false-consensus, effects among groups of high- and low-anxious individuals. In this study, Suls and Wan (1987) identified college students who reported high or low fear toward a variety of objects and situations (e.g., snakes, public speak­ ing, failure). These groups of subjects then were asked to estimate the percent­ age of college undergraduates who had these same fears. The results of this study provided only mixed and relative evidence of self-based consensus. That is, both high and low fear subjects significantly overestimated the incidence of high fear (compared with actual population estimates), and on only three of eight fear items were the estimates of high-fear students greater than those of low-fear students. Future Event Predictions. Two studies (Butler & Mathews, 1983, 1987) ex­ amined anxious individuals’ future event predictions for self and others. The results of both of these studies are quite similar. Because one of these (Butler & Mathews, 1983) employed a group of patients who had received a diagnosis of major depres­ sive disorder, in addition to a group of patients who had received a diagnosis of generalized anxiety disorder and a group of normal control subjects, we focus our discussion on the results of this study. In this study, all subjects were asked to estimate the likelihood that six posi­ tive and six negative events were likely to occur to themselves and to a hypothet­ ical other. Analyses of the subjective probability data revealed no significant effects associated with group assignment or reference condition for the positive events. However, for negative events, there was a groups x reference condition interac­ tion. Both anxious and depressed patients rated negative events as more likely to occur to themselves than did control subjects. In the other-reference condi­ tions, only depressed patients rated negative events as more likely to occur for the hypothetical other compared with control subjects. These results are consis­ tent with the depression evidence reviewed earlier suggesting the existence of a future-event schema (Andersen et al., 1992). Further, these results suggest that anxious patients (like the Andersen et al. mildly depressed, and probably anx­ ious subjects) have not yet, through likely mental rehearsal or rumination about the future, acquired the future-event schema. Memory Processes

Both the Greenberg and Alloy (1989) and Vazquez and Alloy (1987) studies described earlier examined the notion that information consistent with self- and other-schemata would be better recalled (Greenberg & Alloy) or recognized (Vaz­ quez & Alloy) than information irrelevant to the schemata or with information

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

315

not encoded schematically. In this study, nonanxious-nondepressed, depressed, and anxious groups of subjects judged whether lists of positive and negative depression-relevant (DR), anxiety-relevant (AR), or control adjectives described themselves, their best friends (both studies), and people in general (Vazquez & Alloy). Additionally, subjects were asked to indicate whether the adjectives meant the same as some other word. This latter question permitted a comparison of schematic self- and other-encoding versus a semantic encoding of the stimuli. Following this task, subjects were given a 5-minute unexpected recall task (Green­ berg & Alloy) or a surprise recognition test (Vazquez & Alloy). Greenberg and Alloy (1989) analyzed separately subjects’ recall for yes and no responses to the referencing questions. For “yes” responses, all subjects showed enhanced recall for schematically encoded (self or other) versus semantically en­ coded positively valenced adjectives. Also, all subjects recalled more positive and fewer negative adjectives for their best friends than for themselves. For “no” responses, a number of important results emerged. First, there were no differ­ ences for nondepressed-nonanxious, depressed, or anxious groups on recall for control adjectives. However, for the depression-relevant adjectives, both depressed and anxious subjects showed greater recall for schematically than for semanti­ cally encoded stimuli in the self-referent conditions. All groups in the otherreference condition recalled an equal number of schematically encoded and seman­ tically encoded DR adjectives. Finally, for the anxiety-relevant adjectives, only anxious subjects showed greater recall for schematically than for semantically encoded adjectives in the self-reference condition. Again, there were no group differences in recall in the other-referent condition. Thus, these results are con­ sistent with the schema-content findings reported earlier in suggesting that nondepressed-nonanxious, depressed, and anxious subjects’ schemata for others are positive and equally so. The recognition results reported by Vazquez and Alloy (1987) were less in­ formative, because of a ceiling effect (i.e., overall recognition rates were quite high for all subjects and with respect to all types of encoding instructions). The analyses revealed only that adjectives processed schematically with reference to the self were better recognized than those processed schematically with refer­ ence to people in general or those semantically encoded. Recognition rates for adjectives encoded schematically for best friends did not differ from those for any of the other encoding conditions. Summary

The results of studies reviewed in this section suggest that, although the self­ schemata of normal (i.e., nondepressed and nonanxious) individuals are quite positive, the self-schemata for anxious and depressed individuals are more nega­ tive (although both of these latter schemata also contain some positive content; see also Pelham, 1991), the degree of negativity perhaps varying with the level

316

WEARY AND EDWARDS

of dysfunction (at least for depression). However, there was not evidence of cor­ responding group differences in other schemata; normal, depressed, and anxious individuals all appeared to hold equally positive conceptions of others. Indeed, there was only one consistent group difference that emerged for other-referent conditions and only for moderately depressed individuals, namely, the tendency for depressed compared with nondepressed individuals to be more pessimistic in their future-event predictions for others. Although at first this might appear to be evidence of a self-consensus bias, the evidence reviewed suggests that in­ formation about future events is contained in a separate schema that results, in all likelihood, from an increasing tendency to ruminate about the future as the severity or length of depression increases. Predictions of the likely occurrence of positive and negative events for both self and others, then, appear to be in­ fluenced in moderately depressed persons by such future-event schemata, not by self-schemata. What are we to conclude about the absence of depressed-nondepressed and anxious-nonanxious differences in other perception? Are we to conclude that neither depressed, anxious, nor normal individuals use self-schemata, at least in any direct way, as bases for judgments of others? Or should we conclude that normal individuals use their self-schemata as bases for judging others, but depressed and anxious individuals do not? Although we did not focus our discus­ sion on an examination of within depression or anxiety level comparisons of judg­ ments of the self versus others, a number of the studies reviewed permitted such comparisons, and many of these provided evidence (albeit in some cases indirect evidence) that the first conclusion may be the more accurate one (see also Alloy, 1988). That is, the content of other-representations for depressed, anxious, and normal individuals, although equally positive (and perhaps equally accessible), differ from the content of their self-representations.

SOCIAL-COGNITIVE EFFECTS RELATED TO MOTIVATIONAL COMPONENTS OF DEPRESSION AND ANXIETY

Thus far, we have discussed the manner in which the social information process­ ing of depressed and anxious individuals might be influenced by cognitive process­ es and structures that are particular to depression and anxiety. However, as much recent social cognitive research has shown (Higgins & Sorrentino, 1990; Sorrentino & Higgins, 1986), there is much to be gained from considering the ef­ fects of motivational factors on social information processing. This is particularly true in the realm of depression and anxiety, where the underlying motivations may be extreme and the cognitive manifestations unique. There is a large literature on the motivational components of depression and anxiety. Depressed persons typically manifest low self-esteem, feelings of lack

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

317

of control, uncertainty, and a general lethargy; whereas anxious individuals usually show similar feelings of uncontrollability, a strong drive to identify and avoid certain classes of feared stimuli, and a general tendency toward “fight or flight” responses. Although there is a variety of ways to “package” the motivations that are characteristic of the two disorders, we focus on two specific classes of motivations—those related to control and those underlying social comparison processes, particularly self-esteem. We choose this focus because of the long his­ tory of these two concepts in both clinical and social psychology, and because there are well-documented effects of these motivations on other-perception. Control Motivation

Control motivation has been hypothesized to be an important component of both anxiety and depression (Abramson et al., 1989; Beck, 1976; Garber et al., 1980; but see Weary et al., 1993). Moreover, a strong association between depression and less perceived life control, as well as between anxiety and less perceived life control, have been documented repeatedly (Barlow, 1988; Garber, Miller, & Seaman, 1979; Marsh & Weary, 1989; Warren & McEachren, 1983; Weisz, Weiss, Wasserman, & Rintoul, 1987). Evidence regarding the causal role of un­ controllability expectations in depression also has begun to appear recently (see Weary et al., 1993, for a review). Most of the research on the role of control motivation in person perception has employed depressed perceivers. We turn first to an examination of this liter­ ature. Depression. Although there are several studies dealing with the influence of depressives’ expectations of uncontrollability on their social information process­ ing, a model that accounts for these effects has been lacking. Recently, however, Weary et al. (1993) proposed a model that integrates the findings of these studies into a single conceptual framework. Briefly, the model proposes that the effects of mildly depressed individuals’ uncontrollability expectations on their social in­ formation processing are moderated by uncertainty about their ability to under­ stand, predict, and control the social environment. Uncertainty, in turn, leads to the adoption of an accuracy goal, defined, in this case, as the subjective sense that the person understands the social world. This accuracy goal results in exten­ sive processing of and search for social information in an effort to achieve such an understanding and regain control. The effects detailed by this model pertain specifically to mild depression. At more severe levels of depression, this extensive processing of social information would not be expected, because these individuals feel such helplessness that they do not believe that any expenditure of effort will reasonably render the world more understandable or controllable. This model leads to several predictions regarding depressives’ social information

31 8

WEARY AND EDWARDS

processing. First, depressed individuals should be particularly responsive to in­ formation that might render the social environment more understandable, predict­ able, and controllable. That is, depressed subjects should attend to and utilize information that they believe to be diagnostic regarding the social world to a greater extent than nondepressed persons. This prediction has been supported by several studies. McCaul (1983) found that mildly depressed college students were more sensitive than nondepressed students to information regarding the causes of a target person’s actions. Depressed subjects in this study showed less of a correspond­ ence bias than nondepressed subjects when given information about an external causal force. Depressed persons also have been shown to be particularly sensitive to norm violations on the part of a target other (Marsh & Weary, in press; Weary, Jor­ dan, & Hill, 1985). Depressed subjects rate such a person more negatively com­ pared with a normative target than do nondepressed subjects. Further, when depressed subjects are given extra (irrelevant or contradictory) information about the norm violator, their evaluations of the target become even more polarized (Edwards, Weary, & Gleicher, 1991), as might be expected when a person is engaged in greater cognitive activity under the direction of a strong schema (Fiske & Neuberg, 1990; Tesser, 1978). Finally, depressed subjects have been found to recall accurately more details from a behavioral sequence than nondepressed subjects (Edwards et al., 1991). This implies that they attend to such sequences to a greater extent than nondepressives. Although such a finding also might be explained by greater effort at recall, Lassiter and Koenig (1991) provided evidence that depressed subjects tend to uni­ tize (i.e., segment into smaller meaningful units; see Newtson, 1973) schemainconsistent behavioral sequences at a finer rate than do nondepressed subjects. This is consistent with the notion that depressives are paying close attention to potentially diagnostic (i.e., unusual) behavioral sequences. A second prediction made by the Weary et al. model is that depressives ac­ tively will seek out more (diagnostic) social information than will nondepressives. In support of this prediction, Gleicher and Weary (1991, Study 2) found that depressed individuals show greater interest in understanding another person’s per­ sonality and behavior than do nondepressed people. Further, Hildebrand-Saints and Weary (1989) found that depressed subjects were more prone to ask ques­ tions of a target other that were likely to elicit a great deal of information than were nondepressed subjects. Depressed subjects showed this preference regard­ less of the immediate utility of that information. Finally, correlational evidence indicates that depressed individuals are more prone to seek out advice and other social information to cope with daily stressors than are nondepressed persons (Coyne, Aldwin, & Lazarus, 1981). The increased sensitivity to and seeking out of social information on the part of depressives suggests a chronic overanalysis of social information. Hence, a corollary of the first two predictions is that this processing style should have long­

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

3 19

term effects on depressed individuals’ mental structures. Using the Attributional Complexity Scale (Fletcher, Danilovics, Fernandez, Peterson, & Reeder, 1986), several studies have shown that mild depressives tend to have more complex at­ tributional schemata than do nondepressed subjects (Burgess, Baron, & Harvey, 1991; Flett, Pliner, & Blankstein, 1989; Marsh & Weary, 1989). This effect does not extend to severely depressed (i.e., helpless) persons (Marsh & Weary, 1989), a finding that is consistent with the Weary et al. model. The effects cited so far are consistent with the portrait of mild depressives as effortful, vigilant social information processors. However, they do not get directly at the mechanism involved. For instance, it is possible that effects such as these could arise from the use of simple heuristics, an interpretation that is inconsistent with the model. The “effortful processing” paradigm requires more direct measures of cognitive effort. That is, the model predicts that the results of depressives’ processing style should be of a form that is consistent with effort­ ful information processing. Gleicher and Weary (1991) examined the amount and kinds of on-line (Hastie & Park, 1986) inferences generated by depressed and nondepressed perceivers. These researchers identified two types of inferences that might be generated when a person forms an impression: specific inferences, which tend to be tied to in­ dividual behaviors and to be relatively situation specific; and abstract inferences, which tend to be more complex, cross-situational inferences derived from com­ bining various pieces of information (Semin & Fiedler, 1988). Because abstract inferences are formed through the integration of different pieces of information, they should be complex and effortful to construct. Gleicher and Weary predicted and found that depressed subjects generate more overall inferences and a greater number of abstract inferences than do nondepressed perceivers. This was partic­ ularly the case when the target person violates a social norm, as would be ex­ pected if depressed perceivers are particularly sensitive to such information. A second study (Edwards et al., 1991) extended these findings by examining the possibility that depressed perceivers would generate an even greater number of inferences when provided with additional information about the target person. This hypothesis was based on the notion that depressed perceivers, being sensi­ tive to social information, would try to integrate such information into their cog­ nitive representations of the target person. As predicted, depressed subjects were found to generate more overall inferences than nondepressed subjects, especially when given additional information. Although the studies by Gleicher and Weary (1991) and Edwards et al. (1991) offer evidence regarding the content of depressed perceivers’ thoughts, they do not tap directly the process by which depressed and nondepressed perceivers form an impression. The Weary et al. model suggests that depressed perceivers should be prone to use effortful, integrative impression formation processes. Edwards and Weary (1993) provided evidence for this. Utilizing the framework provided by Fiske and Neuberg’s (1990) continuum model of impression formation, they

32 0

WEARY AND EDWARDS

predicted and found that depressed perceivers’ effortful style of social infor­ mation processing leads them to engage in piecemeal or individuating impres­ sion formation processes, even in situations where nondepressed perceivers typically utilize less effortful category-based processing. Consistent with the model, path analysis confirmed that this tendency toward piecemeal process­ ing on the part of depressives was mediated by their feelings of uncontrolla­ bility. Taken together, this group of studies indicates that mildly depressed perceivers’ feelings of uncontrollability lead to a number of social-cognitive effects that are consistent with the notion that depressives tend to use effortful, complex types of information processing. These findings provide strong, converging support for the Weary et al. model, indicating that control motivation plays an important role in the social-cognitive processes of depressed individuals. Anxiety. Although the Weary et al. (1993) model outlined earlier is concerned specifically with the effects of depression on person perception processes, it also is applicable to anxiety. Both expectations of uncontrollability and feelings of uncertainty have theoretically and empirically (Alloy et al., 1990; Garber et al., 1980; Marsh & Weary, in press) been associated with anxiety. Indeed, it has been proposed that pure anxiety results from an uncertain expectation that one is helpless, whereas mixed anxiety-depression results from a certain expectation that one cannot control future outcomes combined with uncertainty about the likely occurrence of negative outcomes. Depression results when this uncertain nega­ tive outcome expectancy becomes certain (i.e., when one is certain that he or she cannot control future outcomes and that negative outcomes are certain to oc­ cur). To the degree that uncertainty about helplessness (uncontrollability) expec­ tations and/or uncertainty about the occurrence of negative outcomes is present, we might expect perceivers generally to process social information in a more vigilant and effortful manner, perhaps in an attempt to regain control and reduce uncertainty. A recent study by Marsh and Weary (in press) provided support for this no­ tion. Specifically, Marsh and Weary suggested that, unlike depression, levels of anxiety should be related linearly to perceivers’ sensitivity to information that might render the social world more understandable and controllable. In particu­ lar, they expected that perceivers with the highest levels of trait anxiety (STAI; Spielberger et al., 1970), but low levels of depression, would show the greatest responsiveness to norm violations on the part of a target other. The results of this study support these predictions. There was a highly signifi­ cant linear relationship between anxiety and sensitivity to the social information. Moreover, highly anxious subjects who also were mildly depressed showed greater sensitivity to the available social information than did highly anxious-highly depressed subjects.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

321

Social Comparison Motives

Since Festinger’s (1954) original statement of social comparison theory, a num­ ber of modifications have appeared. Perhaps one of the more significant is the recognition that social comparison processes often may be driven by motives other than the desire to reduce uncertainty about the accuracy or adequacy of one’s abilities and beliefs; that is, comparison activities also may be guided by self­ enhancement and by self-improvement concerns (Wood, 1989). To our knowledge, the only studies that have focused on social comparison motives and the emotional disorders have employed depressed populations. The results of a number of these studies (Flett, Vredenburg, Pliner, & Krames, 1987; Gibbons, 1986; Rosenblatt & Greenberg, 1988, 1991; Wenzlaff & Prohaska, 1989) provide ample support for the role of self-esteem motivations in compari­ son activities of both depressed and nondepressed subjects. Much of this research has examined the comparison of moods and has found a preference on the part of nondepressives for others who also are nondepressed and on the part of depres­ sives for others who also are depressed. Moreover, some of this research has found that depressed subjects report more positive affect following comparisons with others who feel as bad as or worse than they do. This “downward compari­ son” (Wills, 1981) and resultant positive affect has been viewed as an attempt on the part of depressed individuals to protect or bolster their self-esteem. However, recent evidence reported by Wenzlaff and Prohaska (1989) suggested that this general tendency for depressed persons to prefer worse-off others might be moderated by their attributions for the others’ good or bad mood. In this study, depressed subjects most preferred as a potential partner and felt better after learning about an unhappy other who was not responsible for the events that gave rise to his or her unhappiness, but least preferred and felt worse after learning of another who was responsible for his or her happiness. Nondepressed subjects felt better after learning of another’s happiness and most preferred as a partner the happy person who was not responsible for his or her good mood. In sum­ mary, although depressed and nondepressed subjects had opposite preferences for the mood of their partners, they both preferred people whose emotional states had been determined by external factors. Wenzlaff and Prohaska suggested that, in social comparison activities, “. . . nondepressed individuals seek confirmation for the idea of benevolent fate, whereas depressed people may want to reinforce the view that their unhappiness is the result of misfortune, not personal short­ comings” (p. 231). One final study of comparison preferences should be mentioned. This study (Weary, Marsh, & McCormick, in press), concerned not with mood comparison but with opinion comparison, examined the notion that depressed and nondepressed individuals’ comparison choices may be motivated by different informational motives—construction and validation—as well as self-esteem motives.

322

WEARY AND EDWARDS

Research on the evaluation of opinions and beliefs generally has supported the proposition that individuals prefer to compare themselves to others who are similar on attributes related to the focal judgment. How does comparison with similar others reduce ambiguity about the correctness of individuals’judgments? According to an attribution perspective (Fazio, 1979; Goethals & Darley, 1977), the individual’s task is to decide whether the judgment is entity based (i.e., whether it accurately describes the object of concern) or whether it is person based (i.e., whether it is a biased appraisal of the object caused by the individual’s needs, abilities, or other characteristics). Agreement from a similar other should result in discounting of the entity as a cause of the opinion, and therefore should yield little real information about the veridicality of the judgment. Likewise, disagree­ ment from a dissimilar other should yield relatively little information gain. On the other hand, comparison with a similar other whose judgment disagrees or with a dissimilar other whose judgment agrees with the individual’s should pro­ vide objective information about the adequacy of the judgment, albeit informa­ tion that addresses two different informational concerns. More specifically, the similar disagreer should be chosen when concerns about construction (i.e., concerns that there is some information or understanding about the object that the comparison other possesses that is necessary for an accurate appraisal) are aroused. The dissimilar agreer should be chosen when the desire for validation of a judgment is operating. In this instance, the individual can be more confident that the judgment is an accurate appraisal of the object or en­ vironment, and not due to personal attributes. Weary et al. (in press) predicted that depressed compared with nondepressed subjects would be relatively more uncertain (see Weary, Elbin, & Hill, 1987, for supportive evidence) of their judgments and would be more interested in satis­ fying construction motives; nondepressed subjects were expected to show a prefer­ ence for comparisons that would satisfy validation concerns. Overall, the results were consistent with predictions. Prior to making their comparison selections, depressed subjects tended to express more uncertainty about their judgments of the causes of an event than did nondepressed subjects. Moreover, depressed com­ pared with nondepressed subjects indicated a greater preference for the similar disagreer, whereas nondepressed subjects preferred as a partner the dissimilar agreer to a greater extent than did depressed subjects. The results for several measures designed to tap the informational comparison motives provided addi­ tional support for the notion that subjects’ informational comparison motives de­ termined their comparison choices. Summary

The evidence reviewed in this section indicates that a variety of motives influence depressives’ processing of information about others. Of the two literatures, work on the role of control motivation currently is more extensive and programmatic.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

323

Specifically, the empirical findings reported by Weary and her colleagues pro­ vide compelling evidence that heightened control motivation characteristic of depressives may influence both the representational content and processing of information about others, at least unknown others. Although the content is deter­ mined jointly by social-contextual factors and by depressed perceivers’ observa­ tional goals (uncertainty reduction via subjective accuracy), the process generally appears to be relatively effortful and vigilant. Similarly, depressed individuals’ social comparison activities, at least as indexed by their preferences for informa­ tion about others, appear to be influenced by important motivational concerns. Although the literature is small, there is consistent evidence that self-esteem moti­ vations play an important role in depressives’ and nondepressives’ social compar­ ison choices, albeit with different outcomes for these two groups. There also is some evidence that informational motives may influence differentially depressed and nondepressed individuals’ comparison preferences. The literature on the role of motivation in depressives’ processing of informa­ tion about others is admittedly embryonic. However, the evidence that does exist suggests that, far from being passive recipients of information about the social world, depressed perceivers actively and in a directed fashion construct concep­ tions of others and their actual social environments. The degree to which they are aware of their active participation in construing and constructing their social worlds is an interesting question that must await further research, as are the like­ ly numerous consequences of such motivated cognitive activity. Although we could find only limited research on the role of control motiva­ tion and no research on the role of social comparison motives on anxious per­ ceivers’ processing of information about others, for theoretical reasons we suspect that the findings for mildly and moderately depressed individuals reviewed earli­ er would obtain for anxious individuals; that is, uncertainty reduction and self­ esteem enhancement motives are also characteristic of anxious individuals, and thus should result in similar social perception and social comparison outcomes. This is not to say that there are no motivational differences between moderately depressed and anxious individuals that might affect their social cognitive processes; rather, we only mean to point out that these two sources of motivation are not likely to differentiate these two groups.

CONCLUSION

Perhaps one of the most significant conclusions to be derived from this review of the literature on depression and anxiety is that the self-schema appears not to exert a very powerful or consistent effect on the processing of information about other people. Although this conclusion is consistent with some recent research (Prentice, 1990) on basic self- and other-perception, it is inconsistent with other evidence that suggests a role for the self in other-perception, as well

324

WEARY AND EDWARDS

as with numerous theories (including theories of depression and anxiety) that have emphasized the relationship between the self-concept and the perceptions and evaluations of others. Could it be that self-representation effects would have been more widespread and obvious if researchers had employed more severely depressed and anxious populations, who presumably would have more accessible, better organized, and more affectively consistent self-representations? Perhaps, although the research reviewed in this chapter uncovered no differences between studies that employed subclinical and clinical populations; however, studies employing the latter were few in number. Could it be that the paradigms employed in the research reviewed were ill suited to detect self-representation effects? Did they employ stimulus materials or judgment tasks that did not require perceivers to go beyond the data given (Catrambone & Markus, 1987)? Although this might be a partial answer, it seems unlikely to account for the bulk of the research findings. To be sure, all of the memory studies employed immediate recall measures where we often would not expect to see generally robust self-representation effects in memory for others’ characteristics. However, we also saw little impact of the self on the content of general beliefs about others or on a variety of inferential judgment tasks. Moreover, in most studies, the target other was an unknown other, a hypotheti­ cal other, or a general person type (e.g., average college student, people in gener­ al). We would expect self-representation effects to be most pronounced when the target information is so impoverished (Catrambone & Markus, 1987). Thus, it would seem, on the basis of existing evidence, that the role of depressive and anxious self-schemata, as well as the self-schemata of normal individuals, in so­ cial perception is far from strong or pervasive. Despite this evidence, for several reasons it would seem premature to con­ clude that the normal, depressed, or anxious self-representations play no role in the perception of others. First, as noted, past research has shown that the self, under some conditions and on some measures, plays a role in other-perception. Second, researchers in the studies reviewed here have employed rather limited and perhaps reactive assessments of the content of self- and other-representations. For example, many studies have focused on differences in the traits and evalua­ tive beliefs used to characterize self and others. Although traits and evaluations are prevalent contents of self- and other-representations, basic social cognition research (Fiske & Cox, 1979; Prentice, 1990) has uncovered a number of addi­ tional types of information (e.g., demographic, physical, behavioral, interper­ sonal, covert reactions) contained in such representations, some of which are differentially important in self- and other-representations. For example, Prentice (1990) obtained evidence that, for normal individuals, information about physi­ cal attributes and interpersonal relations are more prevalent in free descriptions of others than of self, as are evaluative judgments. However, privileged in­ formation (i.e., covert reactions) is more prevalent in self-descriptions. Only

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

32 5

information about traits and behaviors appear to be represented equally in selfand other-representations. Studies that employed less reactive, more complete indices of the content of self- and other-representations might have uncovered important domains where the self plays a major role in the content and organiza­ tion of schemata of other people. Third, the experimental paradigms employed in the research reviewed in this chapter have not captured the way that people usually learn and use information about the self and others. That is, such information often is encountered in a so­ cial context, wherein people explicitly are offering information about themselves and others and often are not required to use this information for recall or interpre­ tive purposes for some time (Kahan & Johnson, 1992). There is at least some evidence that people code and recall information according to both referent and source, and that self-schemata play a role in one, but not the other, information and judgment process. Such a focus on the representational effects of the social exchange of personal information might shed light on the social worlds of depressed and anxious individuals. Finally, whatever the verdict on the role of self on depressed and anxious in­ dividuals’ cognitive representations of others, future research would do well to consider the role of other types of representations in such individuals’ social per­ ception processes. For example, there is work indicating that the cognitive representations of familiar, significant others (Andersen & Cole, 1990; Prentice, 1990) differs from those of unfamiliar others and, in some ways, from the self. Moreover, these representations appear to operate schematically, thus serving as effective encoding, organizing, and retrieval aids, as well as sources of in­ ferential errors about others. The research on motivational influences on depressed and anxious individu­ als’ perception of others reviewed earlier highlighted the active role of such per­ ceivers in the construction and construal of their social worlds. It seems likely that the significant-other representations of depressed and anxious individuals differ in important ways from those of nondepressed, nonanxious individuals. Indeed, the social comparison research reviewed here provides important suggestions as to how the significant others might differ for these groups of individuals. In sum, a more complete consideration of the person categories used by depressed and anxious perceivers in understanding others would seem an essential and, to our minds, exciting direction for future research.

EPILOGUE

We began this chapter by charting the cognitive movements within clinical and social psychology, and then quickly turned our attention to a discussion of a sub­ set of the social-cognitive antecedents, correlates, and sequelae of the emotional disorders. It seems necessary to return, albeit briefly, to the more general topic

32 6

WEARY AND EDWARDS

of social cognition and clinical psychology. Where has the convergence of social cognition and clinical psychology taken us during the last decade that, say, the previously predominant psychodynamic and then behavioral, stimulus-response approaches to clinical phenomena had not? What advantages might we have reaped from this integration? We strongly suspect that readers who have stayed with us through this rather long chapter can, at this point, tolerate only brief mention of a few advantages, and so that is what we provide. First, the social-cognitive approach has entailed a different conceptual level of analysis of a variety of clinical phenomena and, in so doing, has permitted significant gains in the conceptualization, assessment, and treatment of psychological disorders. Perhaps nowhere is this more obvious than in the treatment of the emotional disorders where cognitive therapies have been shown to be as effective as pharmacotherapies. Second, the integration of social cognition and clinical psychology has brought with it a new set of empiri­ cal methods for investigating the important effects of basic cognitive mechanisms on adaptive and maladaptive behavioral and affective responses. Third, the in­ tegration has provided important links to other substantive fields of psychology, thereby furthering the hope of many for a more unified, comprehensive under­ standing of human behavior. Future research undoubtedly will establish limita­ tions of this approach. However, it is difficult to believe that clinical psychology will ever again eschew consideration of the important role of mental contents and processes, nor of social psychological factors in the development, assess­ ment, and treatment of the psychological disorders. ACKNOWLEDGMENTS

We thank all of the individuals who so graciously sent their work to us for possi­ ble coverage in this chapter. With your help, we learned a great deal about the high quality and diverse coverage of research programs at the interface of social cognition and clinical psychology. Regrettably, we were unable to cover in any detail more than a handful of the provocative papers sent to us. We also thank the editors of this volume, Bob Wyer and Thom Srull, as well as Faith Gleicher, Kerry L. Marsh, Bill von Hippel, and John Yost for their helpful comments on earlier drafts of this chapter. REFERENCES Abramson, L. Y., Metalsky, G. I., & Alloy, L. B. (1989). Hopelessness depression: A theory-based subtype of depression. Psychological Review, 96, 358-372. Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. (1978). Learned helplessness in humans: Cri­ tique and reformulation. Journal o f Abnormal Psychology, 87, 49-74. Ackermann, R., & DeRubeis, R. J. (1991). Is depressive realism real? Clinical Psychology Review, 11, 565-584.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

327

Alloy, L. B. (Ed.). (1988). Cognitive processes in depression. New York: Guilford. Alloy, L. B., & Abramson, L. Y. (1979). Judgment of contingency in depressed and nondepressed students: Sadder but wiser? Journal o f Experimental Psychology: General, 108 , 441-485. Alloy, L. B., & Abramson, L. Y. (1988). Depressive realism: Four theoretical perspectives. In L. B. Alloy (Ed.), Cognitive processes in depression (pp. 223-265). New York: Guilford. Alloy, L. B., & Ahrens, A. H. (1987). Depression and pessimism for the future: Biased use of statisti­ cally relevant information in predications for self versus others. Journal o f Personality and Social Psychology, 5 2 , 366-378. Alloy, L. B., Hartlage, S., & Abramson, L. Y. (1988). Testing the cognitive diathesis-stress the­ ories of depression: Issues of research design, conceptualization, and assessment. In L. B. Alloy (Ed.), Cognitive processes in depression (pp. 31-73). New York: Guilford. Alloy, L. B., Kelly, K. A., Mineka, S., & Clements, C. M. (1990). Comorbidity in anxiety and depressive disorders: A helplessness/hopelessness perspective. In J. D. Maser & C. R. Cloninger (Eds.), Comorbidity in anxiety and m ood disorders (pp. 499-543). Washington, DC: American Psychiatric Press. Andersen, S. M ., & Cole, S. W. (1990). “Do I know you?”: The role of significant others in general social perception. Journal o f Personality and Social Psychology, 59, 384-399. Andersen, S. M ., Spielman, L. A., & Bargh, J. A. (1992). Future-event schemas and certainty about the future: Automaticity in depressives’ future-event predictions. Journal o f Personality and So­ cial Psychology, 63, 711-723. Anderson, C. A., & Arnoult, L. H. (1985). Attributional models of depression, loneliness, and shy­ ness. In J. H. Harvey, G. Weary, & M. A. Stanley (Eds.), Attribution: Basic issues and applica­ tions (pp. 235-279). New York: Academic Press. Anderson, C. A., Horowitz, L., & French, R. (1983). Attributional style of lonely and depressed people. Journal o f Personality and Social Psychology, 45, 127-136. Anderson, C. A., & Riger, A. L. (1991). A controllability attributional model: Dimensional and situational interactions in the prediction of depression and loneliness. Social Cognition, 4, 270-292. Anderson, J. R. (1978). Arguments concerning representations for mental imagery. Psychological Review, 85, 249-211.

Andrews, B., & Brewin, C. (1990). Attributions of blame for marital violence: A study of ante­ cedents and consequences. Journal o f M arriage and the Family, 52, 757-767. Averill, J. R. (1988). Disorders of emotion. Journal o f Social and Clinical Psychology, 6, 247-268. Banaji, M. R., & Steele, C. M. (1989). The social cognition of alcohol use. Social Cognition, 7, 137-151. Bandura, A. (1974). Behavior theory and models of man. American Psychologist, 29, 859-869. Bandura, A. (1977). Self-efficacy: Towards a unifying theory of behavior change. Psychological Review, 84, 191-125. Bargh, J. A., & Tota, M. E. (1988). Context-dependent processing in depression: Accessibility of negative constructs with regard for self but not others. Journal o f Personality and Social Psychol­ ogy, 54, 924-939. Barlow, D. H. (1988). Anxiety and its disorders: The nature and treatment o f anxiety and panic. New York: Guilford. Barlow, D. H. (1991). Disorders of emotion. Psychological Inquiry, 2, 97-105. Barsalou, L. W. (1990). On the indistinguishability of exemplar memory and abstraction in category representation. In R. S. Wyer & T. K. Srull (Eds.), Advances in social cognition (Vol. 3, pp. 61-88). Hillsdale, NJ: Lawrence Erlbaum Associates. Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects. New York: Hoeber. Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: International Univer­ sities Press. Beck, A. T ., Brown, G., Steer, R. A., Eidelson, J. I., & Riskind, J. H. (1987). Differentiating anxi­ ety and depression utilizing the Cognitions Checklist. Journal o f Abnormal Psychology, 96, 179-186.

328

WEARY A ND EDWARDS

Beck, A. T ., & Emery, G. (1985). Anxiety disorders and phobias: A cognitive perspective. New York: Basic Books. Beck, A. T ., Epstein, N., Brown, G., & Steer, R. A. (1988). An inventory for measuring clinical anxiety: Psychometric properties. Journal o f Consulting and Clinical Psychology, 56, 893-897. Beck, A. T ., Laude, R., & Bohnert, M. (1974). Ideational components of anxiety neurosis. Archives o f General Psychiatry, 31, 319-325. Beck, A. T ., Rush, A. J., Shaw, B. F., & Emery, G. (1985). Cognitive therapy o f depression. New York: Guilford. Beck, A. T ., Ward, C. H., Mendelson, M ., Mock, J., & Erbaugh, J. (1961). An inventory for meas­ uring depression. Archives o f General Psychiatry, 4, 561-571. Blaney, P. H. (1986). Affect and memory: A review. Psychological Bulletin, 99, 229-296. Blatt, S. J., Quinlan, D. M ., Chevron, E. S., MacDonald, C., & Zuroff, D. (1982). Dependency and self-criticism: Psychological dimensions of depression. Journal o f Consulting and Clinical Psy­ chology, 50, 113-124. Blatt, S. J., Wein, S. J., Chevron, E., & Quinlan, D. M. (1979). Parental representations and depression in normal young adults. Journal o f Abnormal Psychology, 88, 388-397. Bower, G. H., & Gilligan, S. G. (1979). Remembering information related to one’s self. Journal o f Research in Personality, 13, 420-432. Bradley, B. P., Gossop, M ., Brewin, C. R., Phillips, G., & Green, L. (1992). Attributions and relapse in opiate addicts. Journal o f Consulting and Clinical Psychology, 60, 470-472. Brehm, S. S. (1976). The application o f social psychology to clinical practice. Washington, DC: Hemisphere. Brewer, M. B. (1988). A dual-process model of impression formation. In R. S. Wyer & T. K. Srull (Eds.), Advances in social cognition (Vol. 1, pp. 1-36). Hillsdale, NJ: Lawrence Erlbaum Associates. Brewer, W. F., & Nakamura, G. V. (1984). The nature and function of schemas. In R. S. Wyer & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 1, pp. 119-160). Hillsdale, NJ: Lawrence Erlbaum Associates. Brewin, C. R. (1984). Attributions for industrial accidents: Their relationship to rehabilitation out­ come. Journal o f Social and Clinical Psychology, 2, 156-164. Brewin, C. R. (1989). Cognitive change processes in psychotherapy. Psychological Review, 96, 379-394. Brewin, C. R., & Furnham, A. (1986). Attributional versus preattributional variables in self-esteem and depression: A comparison and test of learned helplessness theory. Journal o f Personality and Social Psychology, 50, 1013-1020. Brewin, C. R., MacCarthy, B., Duda, K., & Vaughn, C. E. (1991). Attribution and expressed emo­ tion in the relatives of patients with schizophrenia. Journal o f Abnormal Psychology, 100, 546-554. Brown, G. A., & Harris, T. (1978). Social origins o f depression: A study o f psychiatric disorder in women. New York: Free Press. Burgess, M ., Baron, R. S., & Harvey, J. S. (1991, April). A closer look at depression and attribu­ tional complexity: Three replications suggesting a gender difference. Paper presented at the meet­ ing of the Midwestern Psychological Association, Chicago, IL. Butler, G., & Mathews, A. (1983). Cognitive processes in anxiety. Advances in Behavior Research and Therapy, 5, 51-62. Butler, G., & Mathews, A. (1987). Anticipatory anxiety and risk perception. Cognitive Therapy and Research, 11, 551-565. Campbell, J. D. (1986). Similarity and uniqueness: The effects of attribute type, relevance, and in­ dividual differences in self-esteem and depression. Journal o f Personality and Social Psychology, 50, 281-294. Campbell, J. D., & Fehr, B. (1990). Self-esteem and perception of conveyed impression: Is negative affectivity associated with greater realism? Journal o f Personality and Social Psychology, 58, 122-133.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

329

Cantor, N., Mischel, W ., & Schwartz, J. C. (1982). A prototype analysis of psychological situa­ tions. Cognitive P sychology, 14, 45-77. Carver, C. S., & Scheier, M. F. (1981). Attention and self-regulation: A control theory approach to human behavior. New York: Springer-Verlag. Catrambone, R., & Markus, H. (1987). The role of self-schemas in going beyond the information given. Social Cognition, 5, 349-368. Clark, D. A. (1986). Cognitive-affective interaction: A test of the “specificity” and “generality” hypotheses. Cognitive Therapy and Research, 10, 607-624. Clark, L. A. (1989). The anxiety and depressive disorders: Descriptive psychopathology and differential diagnosis. In P. C. Kendall & D. Watson (Eds.), Anxiety and depression: Distinctive and overlap­ ping features (pp. 83-129). New York: Academic Press. Collins, R. L., Taylor, S. E., & Skokan, L. A. (1990). A better world or a shattered vision? Changes in life perspectives following victimization. Social Cognition, 8, 263-285. Cooper, M. L., Russell, M ., & George, W. H. (1988). Coping, expectancies, and alcohol abuse: A test of social learning formulations. Journal o f Abnormal Psychology, 97, 218-230. Coyne, J. C. (1990). Concepts for understanding marriage and developing techniques of marital therapy: Cognition uber alles? Journal o f Family Psychology, 4, 185-194. Coyne, J. C., Aldwin, C ., & Lazarus, R. S. (1981). Depression and coping in stressful episodes. Journal o f Abnormal Psychology, 90, 439-447. Coyne, J. C., & Gotlib, I. H. (1983). The role of cognition in depression: A critical appraisal. P sy­ chological Bulletin, 94, 472-505. Crandell, C. J., & Chambless, D. L. (1986). The validity of an inventory for measuring depressive thoughts: The Crandell Cognitions Inventory. Behavior Research and Therapy, 3, 291-298. Craske, M. G., & Barlow, D. H. (1991). Contributions of cognitive psychology to assessment and treatment of anxiety. In P. R. Martin (Ed.), Handbook o f behavior therapy and psychological science: An integrative approach (pp. 151-168). New York: Pergamon. Crocker, J., Alloy, L. B., & Kayne, N. T. (1988). Attributional style, depression, and perceptions of consensus for events. Journal o f Personality and Social Psychology, 54, 840-846. Crocker, J., Kayne, N. T ., & Alloy, L. B. (1985). Comparing the self to others in depressed and nondepressed college students: A reply to McCauley. Journal o f Personality and Social Psycholo­ gy, 48, 1579-1583. Cutrona, C. E. (1982). Transition to college: Loneliness and the process of social adjustment. In L. A. Peplau & D. Perlman (Eds.), Loneliness: A sourcebook o f current theory, research, and therapy (pp. 291-309). New York: Wiley. Dalgleish, T ., & Watts, F. N. (1990). Biases of attention and memory in disorders of anxiety and depression. Clinical Psychology Review, 10, 589-604. Dawes, R. B. (1986). Representative thinking in clinical judgment. Clinical Psychology Review, 6, 425-441. Dember, W. N. (1974). Motivation and the cognitive revolution. American Psychologist, 29, 161-168. Derry, P. A., & Kuiper, N. A. (1981). Schematic processing and self-reference in clinical depres­ sion. Journal o f Abnormal Psychology, 90, 286-297. DiClemente, C. C. (1986). Self-efficacy and the addictive behaviors. Journal o f Social and Clinical Psychology, 4, 302-315. Dobson, K. S., & Shaw, B. F. (1987). Specificity and stability of self-referent encoding in clinical depression. Journal o f Abnormal Psychology, 96, 34-40. Duval, S., & Wicklund, R. A. (1972). A theory o f objective self-awareness. New York: Academic Press. Edwards, J. A., & Weary, G. (1993). Depression and the impression-formation continuum: Piecemeal processing despite the availability of category information. Journal o f Personality and Social P sy­ chology, 64, 636-645. Edwards, J. A., Weary, G., & Gleicher, F. (1991, April). Depression and social inferences. Paper presented at the meeting of the Midwestern Psychological Association, Chicago, IL.

330

W EARY AND EDWARDS

Epstein, S. (1992). Coping ability, negative self-evaluation, and overgeneralization: Experiment and theory. Journal o f Personality and Social P sychology, 62, 826-836. Eysenck, M. W ., MacLeod, C., & Mathews, A. (1987). Cognitive functioning in anxiety. Psycho­ logical Research, 49, 189-195. Eysenck, M. W ., & Mathews, A. (1987). Trait anxiety and cognition. In H. J. Eysenck & I. Martin (Eds.), Theoretical foundations o f behavior theory. New York: Plenum. Fazio, R. H. (1979). Motives for social comparison: The construction-validation distinction. Jour­ nal o f Personality and Social Psychology, 37, 1683-1698. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117-140. Fincham, F. D., & Bradbury, T. N. (1990a). To arrive where we began: A reappraisal of cognition in marriage and in marital therapy. Journal o f Family Psychology, 4, 167-184. Fincham, F. D., & Bradbury, T. N. (1990b). Psychology and the study of marriage. In F. D. Fin­ cham & T. N. Bradbury (Eds.), The psychology o f marriage: Basic issues and applications (pp. 118-149). New York: Guilford. Fincham, J. D ., & Bradbury, T. N. (1991a). Social cognition: Implications for behavioral assess­ ment and behavior therapy. In P. Martin (Ed.), Handbook o f behavior therapy and psychological science: An integrative approach (pp. 283-307). New York: Pergamon. Fincham, F. D., & Bradbury, T. N. (1991b). Marital conflict: Toward a more complete integration of research and treatment. In J. Vincent (Ed.), Advances in fam ily intervention, assessment and theory (Vol. 5, pp. 1-24). London: Kingsely. Firth-Cozens, J., & Brewin, C. R. (1988). Attributional change during psychotherapy. British Jour­ nal o f Clinical Psychology, 27, 47-54. Fiske, S. T ., & Cox, M. G. (1979). Person concepts: The effects of target familiarity and descriptive purpose on the process of describing others. Journal o f Personality, 47, 136-161. Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression formation, from category-based to individuating processes: Influences of information and motivation on attention and interpreta­ tion. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 1-74). New York: Academic Press. Fiske, S. T., & Taylor, S. E. (1984). Social cognition. Reading, MA: Addison-Wesley. Fletcher, G. J. O., Danilovics, P., Fernandez, G., Peterson, D., & Reeder, G. D. (1986). Attribu­ tional complexity: An individual differences measure. Journal o f Personality and Social Psychol­ ogy, 51, 875-884. Flett, G. L., Pliner, P., & Blankstein, K. R. (1989). Depression and components of attributional complexity. Journal o f Personality and Social Psychology, 56, 757-764. Flett, G. L., Vredenburg, K., Pliner, P., & Krames, L. (1987). Depression and social comparison information-seeking. Journal o f Social Behavior and Personality, 2, 473-484. Foersterling, F. (1985). Attributional retraining: A review. Psychological Bulletin, 98, 495-512. Folkman, S., Lazarus, R. S., Dunkel-Schetter, C., Delongis, A., & Gruen, R. (1986). Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter outcomes. Journal o f Person­ ality and Social Psychology, 50, 992-1003. Fong, G. T ., & Markus, H. (1982). Self-schemas and judgments about others. Social Cognition, 1, 191-204. Foreyt, J. P., & Rathjen, D. P. (1978). Cognitive behavior therapy. New York: Plenum. Furnham, A., & Brewin, C. R. (1988). Social comparison and depression. Journal o f Genetic P sy­ chology, 149, 191-198. Garber, J., & Hollon, S. D. (1980). Universal versus personal helplessness in depression: Belief in uncontrollability or incompetence. Journal o f Abnormal Psychology, 89, 56-66. Garber, J., Miller, S. M ., & Abramson, L. Y. (1980). On the distinction between anxiety states and depression: Perceived control, certainty, and probability of goal attainment. In J. Garber & M. E. P. Seligman (Eds.), Human helplessness: Theory and applications (pp. 131-171). New York: Academic Press.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

331

Garber, J., Miller, W. R., & Seaman, S. F. (1979). Learned helplessness, stress, and the depressive disorders. In R. A. Depue (Ed.), The psychobiology o f the depressive disorders: Implications fo r the effects o f stress (pp. 335-364). New York: Academic Press. Gibbons, F. X. (1985). Social comparison among mentally retarded persons. American Journal o f Mental Deficiency, 90, 98-106. Gibbons, F. X. (1986). Social Comparison and depression: Company’s effect on misery. Journal o f Personality and Social Psychology, 51, 140-148. Gleicher, F., & Weary, G. (1991). The effect of depression on the quantity and quality of social inferences. Journal o f Personality and Social Psychology, 61, 105-114. Goethals, G. R., & Darley, J. M. (1977). Social comparison theory: An attributional approach. In J. M. Suls & R. M. Miller (Eds.), Social comparison processes: Theoretical and empirical p e r­ spectives (pp. 259-278). Washington, DC: Hemisphere. Gotlib, I. H. (1984). Depression and general psychopathology in university students. Journal o f Ab­ normal Psychology, 93, 19-30. Gotlib, I. H ., & McCabe, S. B. (1991). An information-processing approach to the study of cogni­ tive functioning in depression. In E. F. Walker, B. A. Comblatt, & R. H. Dworkin (Eds.), Progress in experimental personality and psychopathology research (Vol. 1, pp. 131-161). New York: Springer. Gotlib, I. H ., & Meltzer, S. J. (1987). Depression and the perception of social skill in dyadic inter­ action. Cognitive Therapy and Research, 11, 41-54. Greenberg, M. S., & Alloy, L. B. (1989). Depression versus anxiety: Processing of self- and otherreferent information. Cognition and Emotion, 3, 207-223. Greenberg, M. S., & Beck, A. T. (1989). Depression versus anxiety: A test of the content specificity hypothesis. Journal o f Abnormal Psychology, 98, 9-13. Greenwald, A. G., & Pratkanis, A. R. (1984). The self. In R. S. Wyer & T. K. Srull (Eds.), Hand­ book o f social cognition (Vol. 3, pp. 129-178). Hillsdale, NJ: Lawrence Erlbaum Associates. Haaga, D. A. F., & Davison, G. C. (1991). Disappearing differences do not always reflect healthy integration: An analysis of cognitive therapy and rational-emotive therapy. Journal o f Psychother­ apy Integration, 1, 287-303. Haaga, D. A. F., Dyck, M. J., & Ernst, D. (1991). Empirical status of cognitive theory of depres­ sion. Psychological Bulletin, 110, 215-236. Hamilton, D. L., Katz, L. B., & Leirer, V. O. (1980). Organizational processes in impression for­ mation. In R. Hastie, T. M. Ostrom, E. B. Ebbesen, R. S. Wyer, D. L. Hamilton, & D. E. Carl­ ston (Eds.), Person memory: The cognitive basis o f social perception (pp. 121-153). Hillsdale, NJ: Lawrence Erlbaum Associates. Hammen, C., Marks, T ., deMayo, R., & Mayol, R. (1985). Self-schemas and risk for depression: A prospective study. Journal o f Personality and Social Psychology, 49, 1147-1159. Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the judgment task is memory-based or on-line. Psychological Review, 93, 258-268. Heider, F. (1958). The psychology o f interpersonal relations. New York: Wiley. Hibbert, G. A. (1984). Ideational components of anxiety: Their origin and content. British Journal o f Psychiatry, 144, 618-624. Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Review, 94, 319-340. Higgins, E. T ., Bond, R., Klein, R., & Strauman, T. J. (1986). Self-discrepancies and emotional vulnerability: How magnitude, accessibility, and type of discrepancy influence affect. Journal o f Personality and Social Psychology, 41, 1-15. Higgins, E. T ., King, G. A., & Mavin, G. H. (1982). Individual construct accessibility and subjec­ tive impressions and recall. Journal o f Personality and Social Psychology, 43, 35-47. Higgins, E. T ., & Sorrentino, R. M. (1990). Handbook o f motivation and cognition: Foundations o f social behavior (Vol. 2). New York: Guilford. Hildebrand-Saints, L., & Weary, G. (1988). Depression and social information gathering. Personal­ ity and Social Psychology Bulletin, 15, 150-160.

332

W EARY AND EDWARDS

Hill, M. G., & Weary, G. (1983). Perspectives on the Journal o f Abnormal and Social P sychology : How it began and how it was transformed. Journal o f Social and Clinical P sychology, 7, 4-14. Hoehn-Hyde, D., Schlottman, R. S., & Rush, A. J. (1982). Perception of social interactions in depressed psychiatric patients. Journal o f Social and Clinical Psychology, 50, 209-212. Hollon, S. D., & Kendall, P. C. (1980). Cognitive self-statements in depression: Development of an automatic thoughts questionnaire. Cognitive Therapy and Research, 4, 383-395. Holloway, W ., & McNally, R. J. (1987). Effects of anxiety sensitivity on the response to hyperven­ tilation. Journal o f Abnormal Psychology, 96, 330-334. Holtzworth-Monroe, A., & Jacobson, N. S. (1985). Causal attributions of married couples: When do they search for causes? What do they conclude when they do? Journal o f Personality and Social Psychology, 48, 1398-1412. Holyoak, K. J., & Gordon, P. C. (1983). Social reference points. Journal o f Personality and Social Psychology, 44, 881-887. Hope, D. A., Gansler, D. A., & Heimberg, R. G. (1989). Attentional focus and causal attributions in social phobia: Implications from social psychology. Clinical Psychology Review, 9, 49-60. Horowitz, M. J. (Ed.). (1991). Person schemas and maladaptive interpersonal patterns. Chicago: University of Chicago Press. Hull, J. G., & Bond, C. F., Jr. (1986). Social and behavioral consequences of alcohol consumption and expectancy: A meta-analysis. Psychological Bulletin, 99, 347-360. Hull, J. G., & Young, R. D. (1983). Self-consciousness, self-esteem, and success-failure as deter­ minants of alcohol consumption in male social drinkers. Journal o f Personality and Social Psy­ chology, 44, 1097-1109. Ingram, R. E. (1986). Information processing approaches to clinical psychology. Orlando, FL: Aca­ demic Press. Ingram, R. E. (1989). Affective confounds in social-cognitive research. Journal o f Personality and Social Psychology, 57, 715-722. Ingram, R. E. (1990a). Depressive cognition: Models, mechanisms, and methods. In R. E. Ingram (Ed.), Contemporary psychological approaches to depression (pp. 169-195). New York: Plenum. Ingram, R. E. (1990b). Self-focused attention in clinical disorders: Review and a conceptual model. Psychological Bulletin, 107, 156-176. Ingram, R. E., & Kendall, P. C. (1987). The cognitive side of anxiety. Cognitive Therapy and Research, 11, 523-536. Ingram, R. E., & Reed, M. J. (1986). Information encoding and retrieval processes in depression: Findings, issues, and future directions. In R. E. Ingram (Ed.), Information processing approaches to clinical psychology (pp. 131-150). Orlando, FL: Academic Press. Janoff-Bulman, R. (1979). Characterological versus behavioral self-blame. Journal o f Personality and Social Psychology, 37, 1798-1809. Janoff-Bulman, R., & Lang-Gunn, L. (1988). Coping with disease and accidents: The role of self­ blame attributions. In L. Y. Abramson (Ed.), Social-personality inference in clinical psychology (pp. 116-147). New York: Guilford. Jordan, J. S., Harvey, J. H., & Weary, G. (1988). Attributional biases in clinical decision making. In D. C. Turk & P. Salovey (Eds.), Reasoning, inference, & judgm ent in clinical psychology (pp. 90-106). New York: Free Press. Kahan, T. L., & Johnson, M. K. (1992). Self effects in memory for person information. Social Cog­ nition, 10, 30-50. Kayne, N. T ., & Alloy, L. B. (1988). Interaction of clinicians’ and patients’ cognitive biases. In L. Y. Abramson (Ed.), Social cognition and clinical psychology (pp. 295-366). New York: Guilford. Keenan, J. M ., & Baillet, S. D. (1980). Memory for personally and socially significant events. In R. S. Nickerson (Ed.), Attention and performance (Vol. 8, pp. 651-669). Hillsdale, NJ: Lawrence Erlbaum Associates. Kelley, H. H. (1972). Causal schemata and the attribution process. In E. E. Jones, D. E. Kanouse, H. H. Kelley, R. E. Nisbett, S. Valins, & B. Weiner (Eds.), Attribution: Perceiving the causes o f behavior (pp. 151-174). Morristown, NJ: General Learning Press.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

333

Kendall, P. C., & Ingram, R. E. (1987). The future for cognitive assessment of anxiety: Let’s get specific. In L. Michelson & M. Ascher (Eds.), Cognitive-behavioral assessment and treatment o f anxiety disorders (pp. 89-104). New York: Guilford. Kendall, P. C., Vitousek, K. B., & Kane, M. (1991). Thought and action in psychotherapy: Cognitivebehavioral approaches. In M. Hersen, A. E. Kazdin, & A. S. Bellack (Eds.), The clinical psychol­ ogy handbook (pp. 596-626). New York: Pergamon. Kendall, P. C., & Watson, D. (1989). Anxiety and depression: Distinctive and overlapping featu res. New York: Academic Press. Kenny, D. A., & Albright, L. (1987). Accuracy in interpersonal perception: A social relations anal­ ysis. Psychological Bulletin, 102 , 390-402. Kruglanski, A. W. (1989). The psychology of being “Right”: The problem of accuracy in social per­ ception and cognition. Psychological Bulletin, 106 , 395-409. Kuiper, N. A., Derry, P. A., & MacDonald, M. R. (1982). Self-reference and person perception in depression. In G. Weary & H. Mirels (Eds.), Integrations o f clinical and social psychology (pp. 79-103). New York: Oxford. Kuiper, N. A., & MacDonald, M. R. (1982). Self and other perception in mild depression. Social Cognition, 1, 223-229. Kuiper, N. A., & MacDonald, M. R. (1983). Schematic processing in depression: The self-based consensus bias. Cognitive Therapy and Research, 7, 469-484. Kuiper, N. A., & Olinger, L. J. (1986). Dysfunctional attitudes and a self-worth contingency model of depression. In P. C. Kendall (Ed.), Advances in cognitive-behavioral research and therapy (Vol. 5, pp. 115-142). New York: Academic Press. Kuiper, N. A., Olinger, L. J., & MacDonald, M. R. (1988). Vulnerability and episodic cognitions in a self-worth contingency model of depression. In L. B. Alloy (Ed.), Cognitive processes in depres­ sion (pp. 289-309). New York: Guilford. Lassiter, G. D., & Koenig, L. J. (1991). Depression and unitization o f behavior. Unpublished manuscript. Leary, M. R., & Miller, R. S. (1986). Social psychology and dysfunctional behavior: Origins, diag­ nosis, and treatment. New York: Springer-Verlag. Leary, M. R., & Schlenker, B. R. (1981). The social psychology of shyness: A self-presentational model. In J. T. Tedeschi (Ed.), Impression management theory and social psychological research (pp. 335-358). New York: Academic Press. Lefcourt, H. M ., Martin, M ., & Saleh, W. E. (1984). Locus of control and social support: Interac­ tive moderators of stress. Journal o f Personality and Social Psychology, 47, 378-389. Lemon, N., & Warren, N. (1974). Salience, centrality, and self-relevance of traits in construing others. British Journal o f Social and Clinical Psychology, 13, 119-124. Lewicki, P. (1983). Self-image bias in person perception. Journal o f Personality and Social Psychol­ ogy, 45, 384-393. Lewinsohn, P. M ., Hoberman, H., Teri, L., & Hautzinger, M. (1985). An integrative theory of depression. In S. Reiss & R. Bootzin (Eds.), Theoretical issues in behavior therapy (pp. 331-359). New York: Academic Press. Linville, P. W. (1987). Self-complexity as a cognitive buffer against stress-related illness and depression. Journal o f Personality and Social Psychology, 52, 663-676. Litz, B. T ., & Keane, T. M. (1989). Information processing in anxiety disorders: Application to the understanding of post-traumatic stress disorder. Clinical Psychology Review, 9, 243-257. Logan, G. S. (1988). Towards an instance theory of automatization. Psychological Review, 9 5 , 492-527. Lyon, D., & Greenberg, J. (1991). Evidence of codependency in women with an alcoholic parent: Helping out Mr. Wrong. Journal o f Personality and Social Psychology, 61, 435-439. MacLeod, C., & Mathews, A. (1991). Cognitive-experimental approaches to the emotional disor­ ders. In P. R. Martin (Ed.), Handbook o f behavior therapy and psychological science: An integra­ tive approach (pp. 116-150). New York: Pergamon.

334

WEARY AND EDWARDS

Maddux, J. E. (1991). Self-efficacy. In C. R. Snyder & D. Forsyth (Eds.), Handbook o f social and clinical psychology: The health perspective (pp. 57-78). New York: Pergamon. Mahoney, M. J. (1974). Cognition and behavior modification. Cambridge, MA: Ballinger. Markus, H. (1977). Self-schemata and processing information about the self. Journal o f Personality and Social Psychology, 35, 63-78. Markus, H ., & Smith, J. (1981). The influence of self-schemas on the perceptions of others. In N. Cantor & J. F. Kihlstrom (Eds.), Personality, cognition, and social interaction (pp. 233-262). Hillsdale, NJ: Lawrence Erlbaum Associates. Marsh, K. L., & Weary, G. (1989). Depression and attributional complexity. Personality and Social Psychology Bulletin, 15, 325-336. Marsh, K. L., & Weary, G. (in press). Severity of depression and responsiveness to attributional information. Journal o f Social and Clinical Psychology. Martin, D. J., Abramson, L. Y., & Alloy, L. B. (1984). The illusion of control for self and others in depressed and nondepressed college students. Journal o f Personality and Social Psychology, 46, 125-136. Maser, J. D., & Cloninger, C. R. (1990). Comorbidity o f m ood and anxiety disorders. Washington, DC: American Psychiatric Press. Mathews, A. (1990). Why worry? The cognitive function of anxiety. Behavior Research and Thera­ p y, 28, 455-468. Mathews, A., Mogg, K., May, J., & Eysenck, M. (1989). Implicit and explicit memory bias in anxiety. Journal o f Abnormal Psychology, 98, 236-240. McCaul, K. D. (1983). Observer attributions of depressed students. Personality and Social Psychol­ ogy Bulletin, 9, 74-82. Merluzzi, T. V., Rudy, T. E., & Krejci, M. J. (1986). Social skill and anxiety: Information process­ ing perspectives. In R. E. Ingram (Ed.), Information processing approaches to clinical psychology (pp. 109-131). Orlando, FL: Academic Press. Metalsky, G. I., Abramson, L. Y., Seligman, M. E. P., Semmel, A., & Peterson, C. (1982). At­ tributional styles and life events in the classroom: Vulnerability and invulnerability to depressive mood reactions. Journal o f Personality and Social Psychology, 43, 612-617. Mineka, S., & Tomarken, A. J. (1989). The role of cognitive biases in the origins and maintenance of fear and anxiety disorders. In T. Archer & L-G Nilsson (Eds.), Aversion, avoidance, and anxi­ ety: Perspectives on aversively m otivated behavior (pp. 195-221). Hillsdale, NJ: Lawrence Erl­ baum Associates. Mogg, K., Mathews, A., & Weinman, J. (1987). Memory bias in clinical anxiety. Journal o f Abnor­ m al Psychology, 96, 94-98. Monroe, S. M ., & Wade, S. L. (1988). Life events and anxiety disorders. In C. L. Last & M. Hersen (Eds.), Handbook o f anxiety disorders (pp. 293-305). New York: Guilford. Moretti, M. M ., & Miller, D. T. (1988). Processing positive and negative information in depres­ sion. Cited in Moretti, M. M., & Shaw, B. F. (1989). Automatic and dysfunctional cognitive process­ es in depression. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 383-421). New York: Guilford. Moretti, M. M ., & Shaw, B. F. (1989). Automatic and dysfunctional cognitive processes in depres­ sion. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 383-421). New York: Guilford. Musson, R. F., & Alloy, L. B. (1988). Depression and self-directed attention. In L. B. Alloy (Ed.), Cognitive processes in depression (pp. 193-220). New York: Guilford. Newtson, D. (1973). Attribution and the unit of perception of ongoing behavior. Journal o f Person­ ality and Social Psychology, 28, 28-38. Nezu, A. M ., & Nezu, C. M. (1989). Clinical decision making in behavior therapy: A problem ­ solving perspective. Champaign, IL: Research Press. Oatley, K., & Bolton, W. (1985). A social-cognitive theory of depression in reaction to life events. Psychological Review, 92, 372-388.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

33 5

Ostrom, T. M. (1984). The sovereignty of social cognition. In R. S. Wyer & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 1, pp. 1-38). Hillsdale, NJ: Lawrence Erlbaum Associates. Park, B. (1986). A method for studying the development of impressions of real people. Journal o f Personality and Social Psychology, 51, 907-917. Pelham, B. W. (1991). On the benefits of misery: Self-serving biases in the depressive self-concept. Journal o f Personality and Social Psychology, 61, 670-681. Pelham, B. W ., & Swann, W. B. (1989). From self-conceptions to self-worth: The sources and structure of self-esteem. Journal o f Personality and Social Psychology, 57, 672-680. Perloff, L. S. (1983). Perceptions of vulnerability to victimization. Journal o f Social Issues, 39, 41-61. Pietromonaco, P. R., & Markus, H. (1985). The nature of negative thoughts in depression. Journal o f Personality and Social Psychology, 48, 799-807. Pietromonaco, P. R., & Rook, K. S. (1987). Decision style in depression: The contribution of per­ ceived risks versus benefits. Journal o f Personality and Social Psychology, 52, 399-408. Prentice, D. A. (1990). Familiarity and differences in self- and other-representations. Journal o f Per­ sonality and Social Psychology, 59, 369-383. Pyszczynski, T., & Greenberg, J. (1987). Self-regulatory perseveration and the depressive self-focusing style: A self-awareness theory of reactive depression. Psychological Bulletin, 102, 1-17. Pyszczynski, T ., Greenberg, J., Hamilton, J., & Nix, G. (1991). On the relationship between self­ focused attention and psychological disorder: A critical reappraisal. Psychological Bulletin, 110, 538-543. Pyszczynski, T ., Hamilton, J. C ., Greenberg, J., & Becker, S. E. (1991). Self-awareness and psy­ chological disfunction. In C. R. Snyder & D. E. Forsyth (Eds.), Handbook o f social and clinical psychology: The health perspective (pp. 138-157). New York: Pergamon. Pyszczynski, T ., Holt, K., & Greenberg, J. (1987). Depression, self-focused attention, and expec­ tancies for positive and negative future life events for self and others. Journal o f Personality and Social Psychology, 52, 994-1001. Rabinowitz, V. C., Zevon, M. A., & Karuza, J. (1988). Psychotherapy as helping: An attributional analysis. In L. Y. Abramson (Ed.), Social cognition and clinical psychology (pp. 177-203). New York: Guilford. Reeder, G. D. (1985). Implicit relations between disposition and behavior: Effects on dispositional attribution. In J. H. Harvey & G. Weary (Eds.), Attribution: Basic issues and application (pp. 87-116). New York: Academic Press. Rehm, L. P. (1977). A self-control model of depression. Behavior Therapy, 8, 787-804. Rosenblatt, A., & Greenberg, J. (1988). Depression and interpersonal attraction: The role of per­ ceived similarity. Journal o f Personality and Social Psychology, 55, 112-119. Rosenblatt, A., & Greenberg, J. (1991). Examining the world of the depressed: Do depressed people prefer others who are depressed? Journal o f Personality and Social Psychology, 60, 620-629. Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions in the attribution process. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 10, pp. 173-220). New York: Academic Press. Ross, L., Greene, D., & House, P. (1977). The false consensus phenomenon: An attributional bias in self-perception and social perception processes. Journal o f Experimental Social Psychology, 13, 279-301. Roth, D., & Rehm, L. P. (1980). Relationships among self-monitoring processes, memory, and depres­ sion. Cognitive Therapy and Research, 4, 149-157. Ruehlman, L. S., West, S. G., & Pasahow, R. J. (1985). Depression and evaluative schemata. Jour­ nal o f Personality, 53, 46-92. Salovey, P., & Turk, D. C. (1988). Some effects of mood on clinicians’ memory. In D. C. Turk & P. Salovey (Eds.), Reasoning, inference, and judgm ent in clinical psychology (pp. 107-123). New York: Free Press. Salovey, P., & Turk, D. C. (1991). Clinical judgment and decision-making. In C. R. Snyder & D. R. Forsyth (Eds.), Handbook o f social and clinical psychology (pp. 416-437). New York: Pergamon.

33 6

WEARY AND EDWARDS

Schultz, R., & Decker, S. (1985). Long-term adjustment to physical disability: The role of social support, perceived control, and self blame. Journal o f Personality and Social Psychology, 48, 1162-1172. Schwartz, R. M ., & Garamoni, G. L. (1989). Cognitive balance and psychopathology: Evaluation of an information processing model of positive and negative states of mind. Clinical Psychology Review, 9, 271-294. Segal, Z. V. (1988). Appraisal of the self-schema construct in cognitive models of depression. P sy­ chological Bulletin, 103, 147-162. Segal, Z. V., Hood, J. E., Shaw, B. F., & Higgins, E. T. (1988). A structural analysis of the self­ schema construct in major depression. Cognitive Therapy and Research, 12, 471-485. Segal, Z. V., & Shaw, B. F. (1986). Cognition in depression: A reappraisal of Coyne and Gotlib’s critique. Cognitive Therapy and Research, 10, 779-793. Semin, G., & Fiedler, K. (1988). The cognitive functions of linguistic categories in describing per­ sons: Social cognition and language. Journal o f Personality and Social Psychology, 54, 558-568. Sewitch, T. S., & Kirtsch, I. (1984). The cognitive content of anxiety: Naturalistic evidence for the predominance of threat related thoughts. Cognitive Therapy and Research, 8, 49-58. Shapiro, J. P. (1988a). Relationships between depressive experience and evaluative beliefs about people in general. Personality and Social Psychology Bulletin, 14, 388-400. Shapiro, J. P. (1988b). Relationships between dimensions of depressive experience and perceptions of the lives of people in general. Journal o f Personality Assessment, 52, 297-308. Shapiro, J. P. (1989). Self-blame versus helplessness in sexually abused children: An attributional analysis with treatment recommendations. Journal o f Social and Clinical Psychology, 8, 442-455. Showers, C. (1992a). Compartmentalization of positive and negative self-knowledge: Keeping bad apples out of the bunch. Journal o f Personality and Social Psychology, 62, 1036-1049. Showers, C. (1992b). Evaluatively integrative thinking about characteristics of the self. Personality and Social Psychology Bulletin, 18, 719-729. Shrauger, J. S., & Patterson, M. B. (1974). Self-evaluation and the selection of dimensions for evalu­ ating others. Journal o f Personality, 42, 569-585. Silver, R. L., Boon, C., & Stones, M. H. (1983). Searching for meaning in misfortune: Making sense of incest. Journal o f Social Issues, 39, 81-101. Singer, J. L., & Salovey, P. (1991). Organized knowledge structures and personality: Person sche­ mas, self schemas, prototypes, and scripts. In M. J. Horowitz (Ed.), Person schemas and maladaptive interpersonal patterns (pp. 33-79). Chicago: University of Chicago Press. Smith, T. W ., & Greenberg, J. (1981). Depression and self-focused attention. M otivation and Emo­ tion, 9, 323-331. Snyder, C. R., & Forsyth, D. R. (1991). Handbook o f social and clinical psychology. New York: Pergamon. Snygg, D., & Combs, A. W. (1959). Individual behavior: A perceptual approach to behavior (rev. ed.). New York: Harper. Solomon, S., Greenberg, J., & Pyszcynski, T. (1991). A terror management theory of social be­ havior: The psychological functions of self-esteem and cultural worldviews. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 93-159). New York: Academic. Sorrentino, R. M ., & Higgins, E. T. (1986). Handbook o f motivation and cognition: Foundations o f social behavior (Vol. 1). New York: Guilford. Spielberger, C. D., Gorsuch, R. L., & Luschene, R. E. (1970). Manual fo r the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press. Spielman, L. A., & Bargh, J. A. (1990). Does the depressive self-schema really exist? In C. D. McCann & N. S. Endler (Eds.), Depression: N ew directions in theory, research, and practice (pp. 113-126). Toronto: Wall & Emerson. Srull, T. K., & Gaelick, L. (1984). General principles and individuals in the self as a habitual refer­ ence point: An examination of self-other judgments of similarity. Social Cognition, 2, 108-121.

6.

SOCIAL COGNITION AND CLINICAL PSYCHOLOGY

337

Stacy, A. W ., Newcomb, M. D., & Bentler, P. M. (1991). Cognitive motivation and drug use: A 9-year longitudinal study. Journal o f Abnormal P sychology, 100 , 502-515. Stacy, A. W ., Newcomb, M. D., & Bentler, P. M. (1993). Cognitive motivations and sensation seeking as long-term predictors of drinking problems. Journal o f Social and Clinical Psychology, 12, 1-24. Stacy, A. W ., Widaman, K. F., & Marlatt, G. A. (1990). Expectancy models of alcohol use. Jour­ nal o f Personality and Social Psychology, 58, 918-928. Stoltenberg, C. D., Pace, T ., & Maddux, J. E. (1986). Cognitive style and counselor credibility: Effects on client endorsement of Rational Emotive Therapy. Cognitive Therapy and Research, 10, 237-243. Strack, S., & Coyne, J. C. (1983). Social confirmation of dysphoria: Shared and private reactions to depression. Journal o f Personality and Social Psychology, 44, 798-806. Straumann, T. J. (1990). Self-guides and emotionally significant childhood memories: A study of retrieval efficiency and incidental negative emotional content. Journal o f Personality and Social Psychology, 59, 869-880. Straumann, T. J. (1992). Nothing ado about much: Overlooked opportunities for cognitive approaches to depression? Psychological Inquiry, 3, 266-269. Stroop, J. R. (1938). Factors affecting speed in serial verbal reactions. Psychological Monographs, 50, 38-48. Suls, J., & Wan, C. K. (1987). In search of the false-uniqueness phenomenon: Fear and estimates of social consensus. Journal o f Personality and Social Psychology, 52, 211-217. Swallow, S. R., & Kuiper, N. A. (1987). The effects of depression and cognitive vulnerability to depression on judgments of similarity between self and other. Motivation and Emotion, 11, 157-167. Swann, W. B. (1984). Quest for accuracy in person perception: A matter of pragmatics. Psychologi­ cal Review, 91, 457-477. Swann, W. B., Hixon, J. G., & De La Ronde, C. (1992). Embracing the bitter “truth”: Negative self-concepts and marital commitment. Psychological Science, 3, 118-121. Swann, W. B., Stein-Seroussi, A., & McNulty, S. E. (1992). Outcasts in a white lie society: The enigmatic worlds of people with negative self-conceptions. Journal o f Personality and Social P sy­ chology, 62, 618-624. Sweeney, P. D., Shaeffer, D., & Golin, S. (1982). Attributions about self and others in depression. Personality and Social Psychology Bulletin, 8, 37-42. Tabachnick, N., Crocker, J., & Alloy, L. B. (1983). Depression, social comparison, and the falseconsensus effect. Journal o f Personality and Social Psychology, 45, 688-699. Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective on mental health. Psychological Bulletin, 103, 193-210. Taylor, S. E., Buunk, B. P., & Aspinwall, L. G. (1990). Social comparison, stress, and coping. Personality and Psychology Bulletin, 16, 74-89. Taylor, S. E., & Crocker, J. (1981). Schematic bases of social information processing. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario symposium on p e r­ sonality and social psychology (Vol. 1, pp. 89-134). Hillsdale, NJ: Lawrence Erlbaum Associates. Taylor, S. E., & Schneider, S. K. (1989). Coping and the simulation of events. Social Cognition, 7, 174-194. Tesser, A. (1978). Self-generated attitude change. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 11, pp. 289-338). New York: Academic Press. Thompson, S. C., & Janigian, A. S. (1988). Life schemes: A framework for understanding the search for meaning. Journal o f Social and Clinical Psychology, 7, 260-280. Turk, D. C., & Salovey, P. (1986). Clinical information processing: Bias inoculation. In R. E. In­ gram (Ed.), Information processing approaches to clinical psychology (pp. 306-324). Orlando, FL: Academic Press. Vazquez, C. V. (1987). Judgment of contingency: Cognitive biases in depressed and nondepressed subjects. Journal o f Personality and Social Psychology, 52, 419-431.

338

WEARY A ND EDWARDS

Vazquez, C. V., & Alloy, L. B. (1987). Schematic memory processes fo r self- and other-referent information in depression versus anxiety. Cited in Greenberg, M. S., Vazquez, C. V., & Alloy, L. B. (1988). Depression versus anxiety: Differences in self- and other-schemata. In L. B. Alloy (Ed.), Cognitive processes in depression (pp. 109-142). New York: Guilford. Vuchinich, R. E ., & Tucker, J. A. (1988). Contributions from behavioral theories of choice to an analysis of alcohol abuse. Journal o f Abnormal Psychology, 97, 181-195. Warren, L. W ., & McEachren, L. (1983). Psychosocial correlates of depressive symptomatology in adult women. Journal o f Abnormal Psychology, 92, 151-160. Weary, G., Elbin, S. D., & Hill, M. G. (1987). Attribution and social comparison processes in depres­ sion. Journal o f Personality and Social Psychology, 52, 605-610. Weary, G., Jordan, J. S., & Hill, M. G. (1985). The attributional norm of internality and depressive sensitivity to social information. Journal o f Personality and Social Psychology, 49, 1283-1293. Weary, G., Marsh, K. L., Gleicher, F., & Edwards, J. A. (1993). Social-cognitive consequences of depression. In G. Weary, F. Gleicher, & K. L. Marsh (Eds.), Control motivation and social cognition (pp. 255-287). New York: Springer-Verlag. Weary, G., Marsh, K. L., & McCormick, L. A. (in press). Depression and social comparison mo­ tives. European Journal o f Social Psychology. Weary, G., & Mirels, H. L. (1982). Integrations o f clinical and social psychology. New York: Ox­ ford University Press. Weary, G., Stanley, M. A., & Harvey, J. H. (1989). Attribution. New York: Springer-Verlag. Wegner, D. M. (1992). You can’t always think what you want: Problems in the suppression of un­ wanted thoughts. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 193-226). San Diego, CA: Academic Press. Weissman, M. M ., & Paykel, E. S. (1974). The depressed woman: A study o f social relationships. Chicago: University of Chicago Press. Weisz, J. R., Weiss, S., Wasserman, A. A., &Rintoul, B. (1987). Control-related beliefs and depres­ sion among clinic-referred children and adolescents. Journal o f Abnormal Psychology, 96, 58-63. Wenzlaff, R. M., & Berman, J. S. (1985, August). Judgmental accuracy in depression. Paper presented at the meeting of the American Psychological Association, Los Angeles. Wenzlaff, R. M ., & Prohaska, M. L. (1989). When misery prefers company: Depression, attribu­ tions, and responses to others’ moods. Journal o f Experimental Social Psychology, 25, 220-233. Williams, C. L., & Poling, J. (1989). An epidemiological perspective on the anxiety and depressive disorders. In P. C. Kendall & D. Watson (Eds.), Anxiety and depression: Distinctive and overlap­ ping features (pp. 317-340). New York: Academic Press. Williams, J. M ., Watts, F., MacLeod, C ., & Mathews, A. (1988). Cognitive psychology and the emotional disorders. New York: Wiley. Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90, 245-271. Wood, J. V. (1989). Theory and research concerning social comparisons of personal attributes. P sy­ chological Bulletin, 106, 231-241. Wood, J. V., Taylor, S. E., & Lichtman, R. R. (1985). Social comparison in adjustment to breast cancer. Journal o f Personality and Social Psychology, 49, 1169-1183. Wortman, C. B., & Brehm, J. W. (1975). Responses to uncontrollable outcomes: An integration of reactance theory and the learned helplessness model. In L. Berkowitz (Ed.), Advances in ex­ perim ental social psychology (Vol. 8, pp. 277-336). Orlando, FL: Academic Press. Wyer, R. S., & Srull, T. K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Lawrence Erlbaum Associates. Zuroff, D. C. (1992). New directions for cognitive models of depression. Psychological Inquiry, 3, 274-277.

7

On the Synergy Between Theory and Application: Social Cognition and Performance Appraisal

Jack M . Feldman

Georgia Institute of Technology

Contents The Domains of Organizational Psychology and Social Cognition: Some General Comments 341 Performance Appraisal 344 The Nature of Job Performance and the Rating Process 345 Cognitive Models of the Appraisal Process 347 A Categorization Model of Appraisal 347 An Information-Seeking Model of Appraisal 350 Summary 351 Accuracy in Performance Appraisal 351 Definitional Issues: Conceptual and Operational 352 Rating “Errors” and Accuracy 355 Construct Validity of Performance Appraisals 357 Halo and Bias in Performance Appraisal 358 Halo 358 Overall Impression Halo 359 Dimensional Similarity Halo 360 Bias 361 Gender 362 Race 365 Other Biases 367 Summary 368 Performance Appraisal Context 369 Evaluative Context 370 Purpose of Observation and Appraisal 371 Organizational and Interpersonal Context 375 Summary 378 Individual Differences 378 Cognitive Complexity 378 Observational Skills 379 Implicit Theories 379 Summary 382 339

34 0

FELDMAN

Training for Performance Appraisal 382 Summary 384 Issues and Directions for Future Research Acknowledgments 386 References 386

384

This chapter has two goals. The first is to present the topic of performance ap­ praisal, a central issue in organizational psychology, from the perspective of so­ cial cognition. The second is to explore the intersection of social cognition and organizational psychology: in particular, what each area of study has to offer the other. This is not to say that performance appraisal is the only area of organizational psychology to which social-cognitive knowledge is applicable, or in which it has been applied. Quite the opposite is true: there is probably no area of organiza­ tional research that has not benefited from the influence of social cognition, or which could not potentially benefit. Likewise, there is probably no area of con­ cern to organizational scholars that could not contribute to theory construction and testing in social cognition, although as of this writing the latter direction of influence has been less frequent. A few relevant areas are: 1. Motivation: the study of processes that energize, direct, and control the persistence of behavior (Kanfer, 1991; Lord & Maher, 1991). 2. Job satisfaction and satisfaction-behavior relationships (Iffaldano & Muchinsky, 1985; Locke, 1976). 3. Negotiation processes (Neale & Bazerman, 1991). 4. Interpersonal relationships and their consequences (Feldman, 1986a; Graen & Scandura, 1986). 5. Job perceptions, their origins, and their relationship to satisfaction and be­ havior (Feldman, 1991). 6. Individual decision making (Feldman & Lindell, 1990; Lopes, 1987). 7. Leadership, in terms of both effectiveness and leader emergence/identifi­ cation (Lord, 1985a; Lord, Foti, & Phillips, 1982). 8. Organizational socialization and social influence processes (Feldman, 1991). 9. Group performance, including decision making and problem solving (Bottger & Yetton, 1988; Isenberg, 1986; Laughlin, 1988). This list is both idiosyncratic and incomplete, as any list must be. The interested reader is referred to recent reviews (e.g., House & Singh, 1987; Ilgen & Klein, 1989), to the Handbook o f Industrial-Organizational Psychology (Dunnette & Hough, 1991), and to research annuals such as Research in Organizational Be­ havior and Research in Personnel and Human Resources Management.

7.

PERFORMANCE APPRAISAL

341

THE DOMAINS OF ORGANIZATIONAL PSYCHOLOGY AND SOCIAL COGNITION: SOME GENERAL COMMENTS Organizational psychology is an applied discipline whose goals are the enhance­ ment of human welfare and organizational functioning. Therefore, organization­ al researchers have focused their efforts on issues, problems, or phenomena relevant to one or both of these goals (e.g., motivational processes and their im­ pact on quality or quantity of job performance; performance judgments by peers, subordinates, and superiors; the nature and origins of job satisfaction and its rela­ tionship to work behavior and other life domains; how superior-subordinate rela­ tionships influence individual and group performance, career progress, etc.). In the investigation of these and other topics, any of a variety of specific theoretical positions might be taken, and a number of levels of analysis, from the demo­ graphic to the genetic, might be employed. This is not to say that research in organizational psychology is theoretically bereft or mindlessly eclectic. Rather, organizational theories tend to be broad­ band approaches, seeking to model the development of a given outcome by cap­ turing all of the relevant influences (e.g., Mobley, Griffeth, Hand, & Meglino’s [1979] model of the voluntary turnover process; Ilgen & Feldman’s [1983] expo­ sition of performance appraisal; or Kanfer & Ackerman’s [1989] model of the effects of goals, individual differences, task difficulty, and training on task per­ formance). The criterion of success is often “How much of the variance in a given behavior or outcome is captured by the model?” or, alternatively, “How well does the model, expressed in structural equation form, fit the pattern of observed rela­ tionships?” Research in organizational psychology is strongly field oriented. Although ex­ perimentation has become increasingly important in the last 15-20 years, survey and other field research strategies still characterize the discipline. Coupled with this orientation is a strong psychometric tradition, a great deal of concern with individual differences (e.g., in ability or personality), and a generally integrationist approach to theory building (see, e.g., Schneider, 1983). In contrast, social cognition (in the words of Sherman, Judd, & Park, 1989) “. . .is not a domain of inquiry within social psychology, but an approach or set of assumptions guiding research in a variety of substantive domains” (p. 281). In other words, it is an approach to developing theory in traditional social psy­ chological areas of interest: person perception, attitude development and change, self-concept, social influence, group process, and so forth. Application is not its primary focus, although individual researchers have pursued specific topics rele­ vant to human welfare (e.g., Cialdini, Kallgren, & Reno, 1991; Devine, 1989). Furthermore, interest is not centered principally on explaining specific phenomena. Rather, the focus is on providing accounts of fundamental processes that may underlay a wide range of judgments and behavior. This focus renders social cog­ nition research both broader and narrower in scope than that of organizational

342

FELDMAN

psychology. It is broader in that the processes studied may be applied to a variety of topic areas. For instance, Higgin’s (1989) self-discrepancy model may be brought to bear usefully on questions of task motivation and self-regulation. However, it is narrower, in that the cognitive processes studied may not pro­ vide a complete account of some specific behavior. For example, studies of the automaticity of priming effects on categorization and inference would not systematically consider the role of social desirability or external constraints on the rating response (except as confounding factors). However, these are of cen­ tral interest in the real-world application of social cognition to performance ap­ praisal. There, implicit or explicit contingencies strongly influence the rating response, if not the judgment itself (Banks & Murphy, 1985; Longnecker, Sims, & Gioia, 1987). Social cognition research is virtually entirely laboratory based; it also has a strong situational focus. Individual differences are examined from the perspec­ tive of their implications for processing, and not from a psychometric viewpoint (see Bargh, 1989, on chronic accessibility; Markus, 1977, on schematicity). This conflicts with organizational psychology’s strong integrationist and dispositional leaning. Furthermore, because the experimental conditions and stimulus materi­ als of the prototypical social cognition study are chosen for their utility in illuminat­ ing the hypothetical processes in question, they often fail to reproduce conditions common in the organizational environment (e.g., multiple tasks, high involve­ ment, outcome dependence, frequent interaction among persons, etc.; see, e.g., Ilgen & Favero, 1985.) Although there is a strong “situational” camp in organiza­ tional psychology (e.g., Davis-Blake & Pfeffer, 1989; Salancik & Pfeffer, 1977), there is a countervailing dispositional perspective as well (e.g., Arvey, Bouchard, Segal & Abraham, 1989; Staw & Ross, 1985). The position taken in this chapter is that the fields of social cognition and or­ ganizational psychology have much to offer one another. Rather than simply im­ porting concepts and methods from social cognition, organizational research provides a particularly fruitful range of social environments and types of inter­ personal behavior in which cognitive processes may be studied. In so doing, it provides a venue in which to answer Neisser’s (1980) call for a more “social” social cognition, expanding the parameters of research well beyond those that may be approached in the laboratory. Social cognition, in turn, can provide a coherent theoretical perspective, or at least a meta-theory, stimulating insights into the processes underlying phenomena of practical importance, for instance, goal-performance relationships (e.g., Kanfer, 1991; Kanfer & Ackerman, 1989; Locke & Latham, 1990), the causes of emergent leadership (e.g., Lord, 1985a), and other classic issues of organizational research. Such understanding may prompt the design of more effective interventions (e.g., Feldman, 1986b, 1992a, on per­ formance appraisal). This chapter is organized around two general themes. The first is a weak ver­ sion of the “constructivist” perspective: that an individual’s social reality is con­

7.

PERFORMANCE APPRAISAL

3 43

structed from environmental inputs of varying strengths, as mediated by existing “cognitive structures and strategies.” These operate in the service of long- and short-term goals (see, e.g., Bargh, 1989; Feldman, 1991; Feldman & Lindell, 1990; Jussim, 1991; Kunda, 1990; Murphy & Medin, 1985; Showers & Cantor, 1985). Three general determinants of the cognitions and affect that drive behavior are assumed: individual cognitive/affective memory structures of varying degrees of elaboration, inputs from and interaction with other people in various forms, and the nonsocial environment (e.g., tasks, physical features of the world, etc.). A pattern of reciprocal causation exists among these, such that each may affect the others in a dynamic fashion (James & Tetrick, 1986). This dynamic construction is influenced by two generic characteristics of any given input (Feldman, 1992b; Feldman & Lynch, 1988). One is accessibility, or the ease with which a given cognition (e.g., cognitive structure, heuristic, pri­ or judgment, affective response, etc.) may be brought to mind. A counterpart to accessibility in the realm of the environment is salience, the prominence of a stimulus (or stimulus attribute) in the perceptual field. Note that salient stimu­ lus attributes may influence the accessibility of cognitions (Taylor & Fiske, 1978), and accessible structures may direct attention to stimulus attributes, effectively rendering them more salient (Wyer & Srull, 1986; see also Bargh, 1989, on prim­ ing and attentional responses). The second characteristic, diagnosticity, is defined as the degree to which a given cognition, behavior, or affective response is perceived to imply the nature of a second judgment, affective response, or behavior. This is a broader defini­ tion than that of Feldman and Lynch (1988), who discussed the measurement context alone. Diagnosticity is sometimes a product of the situation. For exam­ ple, framing a question in terms of choices between uncertain alternatives versus relative evaluations of each of two or more objects influences perceptions of the diagnosticity of outcome probability versus outcome value; whether one is aware of the source of one’s mood influences its perceived diagnosticity for judgments of life satisfaction (Schwarz & Bless, 1991). In other words, prior knowledge and/or individual dispositions may influence the perceived diagnosticity of a pri­ or behavior, stimulus attribute, or affective/emotional response for a subsequent response. It is assumed that a satisficing principle applies to the generation of judgments and behavior. That is, the most accessible cognition sufficient to determine a response is used (Feldman & Lynch, 1988; Srull & Wyer, 1989). Whether a given input is sufficient depends on its diagnosticity in relation to that of potential al­ ternative inputs, and on the standards of judgment applied. This introduces the second theme, one that is common in social cognition: the form and extent of information processing depends on both the motivation of an individual to devote cognitive resources to a problem, and the individual’s capacity (given limited resources) to process information at the time of decision, judgment, or perception.

344

FELDMAN

The more consequential a given judgment is, the more the individual is moti­ vated to search for highly diagnostic information (e.g., via effortful processing strategies; Brewer, 1988; Fiske & Neuberg, 1990; Petty & Cacioppo, 1986). Con­ sequentially may arise via external contingencies or individual value structures (Devine, 1989; Eagly & Chaiken, 1993; Johnson & Eagly, 1989; Petty & Caciop­ po, 1986; Tetlock, 1985). For inconsequential decisions or judgments, the out­ put of simple, perhaps automatic, heuristics or categorization processes may be sufficiently diagnostic. However, if simultaneous tasks or other demands on ca­ pacity prevent extensive processing, less diagnostic, but more accessible, inputs to judgment and behavior might be used even if a decision or judgment is conse­ quential (Devine, 1989; Fazio, 1990; Gilbert, 1989; Petty & Cacioppo, 1986). The trouble with such general themes is that they easily can become circular. Thus, the determinants of accessibility and perceived diagnosticity must be speci­ fied in advance in any given situation, as must the nature of involvement, moti­ vations that determine processing goals, and other stimuli to cognitive effort. Likewise, both task and situational factors that may engage capacity need to be specified, and their relationship to individual differences in capacity (and capaci­ ty utilization, such as expertise) need to be outlined. Much of this necessary work is relegated to the pretest phase of social cognition research, determining, for instance, cues that signal category membership of stimulus persons or the degree of outcome dependence of one person on another. Yet (as is seen), these issues are often at the heart of substantive research in organizational psychology.

PERFORMANCE APPRAISAL

Performance appraisal is a major component of “the criterion problem,” a central issue in industrial and organizational psychology (Austin & Villanova, 1992; Smith, 1976). Basically, the criterion problem refers to the specification and meas­ urement of job or task behavior, and frequently its evaluation on some scale of quality or benefit to the organization. Such measures serve as the criteria for the evaluation of selection strategies, training programs, motivational interventions, and the like, as well as input to a wide variety of personnel decisions. They also serve to support many organizational functions. For instance, Cleveland, Mur­ phy, and Williams (1989) found that four factors describe the uses of appraisal information: 1. Making distinctions between individuals (e.g., salary/promotion/termina­ tion). 2. Making distinctions within individuals (e.g., identifying individual train­ ing needs, feedback, transfer and job assignment decisions). 3. Systems maintenance (e.g., personnel planning, identifying organization­ al training needs, reinforcing organizational authority structures).

7.

PERFORMANCE APPRAISAL

34 5

4. Documentation (e.g., criteria for the validation of selection systems, docu­ mentation for personnel decisions, and for meeting legal requirements). As might be expected, any activity with such wide-ranging interpersonal and organizational impact is the focus of a great deal of political activity within the organization, as well as academic controversy (Longnecker et al., 1987). As is seen, this has implications for judgment processes. The Nature of Job Performance and the Rating Process

Job performance is inherently multidimensional. Even in the most routine jobs, all aspects of the meaning of performance cannot be captured adequately by purely objective or hard criterion measures, such as output, sales volume, or the per­ centage of goods that meet specifications. For instance, factors such as “extra­ role behavior” (e.g., courtesy, cooperativeness; MacKenzie, Podsakoff, & Fet­ ter, 1991) are also considered part of job performance. In managerial, technical, professional, and creative jobs, much of the meaning of performance stems from implicit or explicit values applied to products or behavior, and a person’s work cannot be appraised entirely on the basis of countable criteria without introduc­ ing serious distortion into the process (Feldman, 1986b, 1992a). Therefore, judg­ ment plays a major role in performance appraisal, and these judgments typically are expressed on various forms of rating scales. During most of the history of performance appraisal, research has been guided by questions about the form and content of such rating scales, and how form and content relate to the adequa­ cy of the resulting ratings as indicators of the construct of performance (Landy & Farr, 1983). Briefly, the psychometric tradition of industrial psychology led researchers and practitioners alike to conceptualize appraisal as a measurement problem, solu­ ble within the context of classical measurement theory. It was (implicitly or ex­ plicitly) assumed that job performance was definable via a common set of constructs, and that ratings would reflect performance on each of these constructs if only the rater was provided with appropriate guidance. The rater was assumed to observe behavioral incidents about each employee veridically, and to retain these incidents in memory. At the time of judgment, the incidents were assumed to be recalled and integrated to form a judgment (as guided by scale form and content). These assumptions led naturally to the use of rater error measures as criteria of rating quality (Austin & Villanova, 1992; Landy & Farr, 1980, 1983; Sulsky & Balzer, 1988). For example, if raters base their responses on a veridi­ cal representation of each employee’s behavior and integrate these to form a judg­ ment on each of a set of independent dimensions, judgments along these dimensions should be uncorrelated. Furthermore, performance ratings should be distributed more or less normally around the scale midpoint, and should not be associated with factors that are logically extraneous to performance (e.g., the similarity of rater and ratee).

34 6

FELDMAN

Unfortunately, ratings often failed to meet these criteria. For example, the problem of halo error has been of concern to researchers since Thorndike’s time (Austin & Villanova, 1992; Balzer & Sulsky, 1992; Cooper, 1981). A shift in rating content from traits (e.g., dependability) and other person characteristics (Wexley & Klimoski, 1984) to dimensions of behavior did little to eliminate the problem of rating errors. Behaviorally based or “process” appraisal systems, such as Behaviorally Anchored Rating Scales (BARS), Behavioral Expectation Scales (BES), Behavioral Observation Scales (BOS), and the like (Landy & Farr, 1980, 1983; Smith, 1976), although benefiting from more careful specification of con­ tents and evaluative standards, proved no better than graphic rating scales when similar care was employed in their development (Landy & Farr, 1980, 1983). Part of the problem of halo, leniency, and the like is due to these initial assumptions—that rating dimensions should be relatively independent and that scores should be distributed normally about a scale midpoint. As Hulin (1982), Murphy (1982), and Harvey (1982) pointed out, the nature of human abilities is such that some intercorrelation among them (and the aspects of performance that depend on them) is to be expected. Furthermore, the nature of selection, termination, employee development processes, and social interaction on the job (Feldman, 1986b) is such that one would expect the set of people who have been on the job for any length of time to have developed a joint set of those skills required for job performance, and these skills would be correlated naturally with­ in that subpopulation. Likewise, the distribution of performance of retained em­ ployees normally should be skewed, with the modal value somewhat above any scale midpoint. To an observer expecting ratings to be distributed normally about the scale midpoint (typically labeled average or acceptable), such a distribution would ap­ pear to be lenient. Thus, leniency effects might appear to be a more frequent or severe problem than they really are. However, it is also true that rating errors occur. Cooper (1981) discussed “il­ lusory halo,” or a correlation between ratings that is greater than would be ex­ pected on the basis of the correlation between true performance scores. Aside from rating errors, biased judgments due to performance—extraneous charac­ teristics of the ratee also occur in varying magnitude, as least under certain cir­ cumstances (Eagly, Makhijani, & Klonsky, 1992; Ford, Kraiger, & Schectman, 1986; Kraiger & Ford, 1985; Sackett & Dubois, 1991; but see also Swim, Borgida, Maruyama, & Myers, 1989; Waldman & Avolio, 1991). Effects of rating context (e.g., assimilation-contrast) occur as well (Maurer & Alexander, 1991; Murphy, Gannett, Herr, & Chen, 1986). Despite the demonstrated presence of error in appraisal ratings, substantial accuracy does occur in both field and laboratory settings (Borman, 1978; Bor­ man, White, Pulakos, & Oppler, 1991; Nathan & Alexander, 1985; Smither, Barry, & Reilly, 1989). It is also true that measures of rating error (e.g., halo, leniency) show little if any systematic relationship to accuracy indices. In fact,

7.

PERFORMANCE APPRAISAL

347

they sometimes are related positively to accuracy (Becker & Cardy, 1986; Mur­ phy & Balzer, 1986). The question becomes, then, what processes might account for the simultaneous presence of “error,” “bias,” and accuracy in appraisal judg­ ments, and how might an understanding of these processes be used to improve performance appraisal? These questions have been the focus of appraisal research for more than 15 years.

COGNITIVE MODELS OF THE APPRAISAL PROCESS

Models of the appraisal process with substantial cognitive implications predate the formal consideration of social cognition. Wherry (1952, reprinted in Landy & Farr, 1983) anticipated much of the recent concern with memory organiza­ tion, schematic influences on judgment, and other “classic” social-cognitive is­ sues. However, detailed process models were immediately inspired by attribution theory (Atkin & Conlon, 1977; Bazerman & Atkin, 1978; Mitchell, Green, & Wood, 1981; Stone & Slusher, 1975). Research in the attribution theory tradi­ tion sought to explain performance judgments in terms of causal attributions for observed performance based on immediately available stimulus information. However, these approaches could not explain the presence of both rating error and valid judgments as the outcome of a single process. Nor could they explain how—in the context of multiple tasks, competing demands on the rater’s atten­ tion, the presence of performance-extraneous cues, and substantial time delays between observation and rating—the appraisal process took place in the real world. A Categorization Model of Appraisal

Feldman (1981) provided a comprehensive account of the appraisal process, which was revised and expanded by Ilgen and Feldman (1983) to include interpersonal and organizational factors influencing both judgments and rating responses. DeNisi, Cafferty, and Meglino (1984) presented a model based on both attribution theory and Wyer and Srull’s (1980) representation of person judgments. They focused on the perceiver as an active information gatherer, whereas Feldman (1981) and Ilgen and Feldman (1983) concentrated on automatic and controlled categorization processes and memory-based responses to various appraisal in­ struments. The DeNisi et al. research program focused on appraisal purpose and its consequences for memory organization and accuracy. Both the Feldman and the DeNisi et al. models can be discussed in terms of the constructive process concepts outlined earlier. It can be argued that both are concerned with the ac­ cessibility and diagnosticity of information used to make performance judgments, and with situational and individual difference factors influencing these. Feldman’s model also was concerned with the conditions under which central processing of performance information takes place.

348

FELDMAN

Feldman’s model, which was stimulated by earlier work on stereotyping (Feld­ man, 1972; Feldman & Hilterman, 1975), emphasizes the sources of performance—extraneous bias in appraisals (i.e., due to race, gender, age, etc.). Target persons are said to be categorized via one of two processes, automatic or controlled; the automatic corresponds to Newtson’s (1980) “perceptual” mode, and the controlled to his “inferential.” The automatic process is based on highly overlearned cues and is nonvolitional (Bargh, 1989). It operates in parallel with other processing (as it must, given distracting tasks and other demands on process­ ing capacity). Persons are said to be categorized based on the overlap of their salient features with accessible category representations (in 1981, prototypes; but see Barsalou, 1987). Salience is determined by environmental/contextual factors (e.g., novelty, vividness; Taylor & Fiske, 1978) and by features of accessible category representations. The accessibility of concepts can be chronic (Bargh, 1989), determined by schemata (organized knowledge structures; Brewer & Nakamura, 1984) characteristic of implicit theories (of persons, jobs, etc.), value systems, or expertise in some domain. Alternatively, accessibility can be influenced by transitory factors such as affective state, recent category use, or salient stimu­ lus features (e.g., a “token” individual will elicit attention and accessibility of categories relevant to their token status, such as race). Implicitly, then, environ­ mental salience and cognitive category accessibility are reciprocally causally relat­ ed, each influencing the other. The controlled process, described initially in terms of attribution theory, gener­ ally is elicited by environmental factors such as a discrepancy between expected and observed behavior. This discrepancy motivates attention to the inferential process, promoting conscious consideration of causal factors (i.e., consensus, distinctiveness, consistency) and prompting information search (Gilbert, 1989; Newman & Uleman, 1989). No distinction is made between central and peripheral processing, and biases in attribution are explained in terms of the operation of automatic processes that govern attention and category accessibility (Bargh, 1989). The end result of the controlled process is a categorization, typically in terms of a trait-like construct. Categorization is said to have consequences for information organization in memory, prompting retention of category consistent behaviors (unless inconsisten­ cies prompt attributional processing). It also promotes confirmatory hypothesis testing (Snyder & Swann, 1978) and, perhaps, self-fulfilling prophecies (Jussim, 1986; Snyder, 1984). Confirmatory hypothesis testing is not necessarily a bias; it can be a useful strategy under many circumstances (the “positive test strategy”; Klayman & Ha, 1987). It appears to be a default heuristic, prompted by the sub­ ject’s implicit belief that they are being asked to “prove” the hypothesis presented to them (in the legal sense). Disconfirmatory strategies may be elicited by ap­ propriate question framing and instructions (Feldman & Lindell, 1990; Snyder, 1984). Self-fulfilling prophecies typically are mediated by interpersonal behavior, perhaps nonverbal and not under conscious control (such as voice tone, gaze,

7.

PERFORMANCE APPRAISAL

349

etc.), producing category-consistent behavior in the recipient via similarly un­ conscious processes. Snyder (1984) and Jussim (1986) provide extensive discus­ sion; Dipboye (1985) and Feldman (1986a) examine the concept in work settings. Categorization also promotes affective responses to target persons in a process similar to that postulated by Fiske and Pavelchak (1986). From this perspective, the validity of performance appraisals is influenced by the availability and acces­ sibility of valid, job-relevant category systems, and the more extensive process­ ing promoted by outcome dependence, frequent association, and other influences on motivation to process. Thus, the circumstances under which person impressions initially are formed create impressions that are differentially diagnostic for a variety of subsequent judgments. These impressions largely are formed on-line and spontaneously. The nature of subsequent demands for judgment determines whether retrieval of ex­ isting impressions, inferences from categorical impressions, or retrieval of be­ havior from memory are necessary. Variation in the form and content of appraisal instruments determines the diagnosticity of prior judgments and behavioral memories. Feldman (1981) did not consider relative accessibility, but assumed that integration of recalled behavior only occurs where prior on-line judgments sufficient to guide responding do not exist. For example, he postulated that when a job assignment has to be made, previously formed trait judgments are mapped onto a concept of the job’s “situational prototype,” and persons whose represen­ tations are most similar to the job prototype in question are selected. A formal appraisal that requires an “overall effectiveness” rating will be responded to in terms of a global evaluation, as will multiple scales with substan­ tial evaluative loadings, but no specific, job-relevant reference (e.g., trait rating scales). Alternatively, requests for trait ratings may be responded to in terms of a category representation, if the representation includes trait concepts that are diagnostic for the ratings in question. Both guarantee the presence of halo effects in rating, the first due to a common evaluative source for all ratings, and the second due to beliefs about trait co-occurrences in category representations or implicit theories (Schweder, 1977). Behaviorally based scales also may be respond­ ed to by accessing prior judgments and mapping these onto the dimensions of the instrument. The validity of behavioral scales, like that of graphic scales (which require a rating of the degree of some attribute without behavioral anchoring of each scale point), depends on the match between stored representations and the scales’ contents. This explains why (given equally careful scale development) be­ haviorally based and graphic scales offer equivalent validity. In contrast, forced-choice scales depend on making the most accessible response—the overall evaluation—nondiagnostic by forcing the rater to choose between equally positive or negative response options. This requires the rater to search memory for subsidiary judgments to map onto the scale contents (Feld­ man & Lynch, 1988). In these cases, the validity of the appraisals generated is

35 0

FELDMAN

a function of the validity of the judgments or memories that are retrieved to guide the selection of alternatives. In short, then, Feldman (1981) and Ilgen and Feldman (1983) described influ­ ences on the on-line processing of information that make particular kinds of judg­ ments accessible in memory. These judgments may or may not be diagnostic for subsequent judgments and decisions required by different rating scales or organiza­ tional imperatives. To the extent that they are diagnostic, they are used to generate ratings or choices; to the extent that they are not, retrieval of other judgments and/or episodic memories is required. The validity of appraisals is a function of the validity of judgments made previously, as guided by accessible schemata, and of the effec­ tiveness of retrieval and integration processes, in case new judgments are required. An Information-Seeking Model of Appraisal

A recent cognitive model of the appraisal process (DeNisi et al., 1984) focuses attention on the role of processing objectives for information acquisition, storage, and retrieval, and the implications of these processes for appraisal accuracy, rather than on the categorization process. Also, unlike the Ilgen and Feldman and Feld­ man models, it focuses on controlled processing, on the retrieval of information for reprocessing, and on the perception of multiple target persons. DeNisi et al. did not formally consider extraneous person categorizations (e.g., race, age), nor was the role of automaticity explicitly outlined. Rather, they stressed active, in­ tentional information gathering. However, they did give a prominent role to the influence of appraisal objectives on the accessibility of those categories used to encode information about ratees. Category accessibility, in turn, influences the type of information sought. DeNisi et al. (1984) postulated that observed behavior is encoded in terms of processing goals (and the categories they make accessible), plus any precon­ ceived notions held about the individual. Goal specification is defined in terms of the specific rating scales used and the purpose of the evaluation. Thus, in con­ trast to Feldman’s model, their model assumes an influence of scale form on en­ coding, rather than on retrieval and judgment (although they did stress the importance of valid accessible category systems to accuracy). The impact of ex­ traneous categorizations on the encoding of behaviors, which is central to the Feldman model, is handled through the postulation of preconceived notions. Spe­ cifically, the appraiser is said to actively search for information. This search is guided consciously by the appraiser’s specific objectives. It is limited, however, by external constraints such as time pressure, being focused on the information most diagnostic for the particular goal at hand. The types of information sought are postulated based on Kelley’s (1967) attribution model and the Weiner et al. (1971) stable/unstable dimension of causal inference. That is, people are said to seek information about consensus, distinctiveness, and consistency (the latter two especially) to make attributions about the degree of internality and stability of

7.

PERFORMANCE APPRAISAL

351

the causes of behavior. DeNisi et al. explicitly assumed that causal attribution precedes dispositional judgments. Moreover, unlike Feldman, they did not limit the occurrence of controlled attribution processes to conditions in which spon­ taneous dispositional judgments cannot be made or are contradicted by observed behavior. To this extent, their assumptions run counter to attribution research (Gilbert, 1989; Newman & Uleman, 1989). However, they anticipated recent writing in “motivated cognition” (Kunda, 1990) in their discussion of the role of rating objectives. For instance, they proposed that people are motivated to seek more diagnostic negative information when they anticipate no interpersonal consequences of giving low ratings, but seek positive information when ratings are to be used administratively, and thus are likely to influence a subordinate’s welfare. Summary

In short, the DeNisi et al. (1984) model emphasizes different aspects of the ap­ praisal process than did Feldman (1981) and Ilgen and Feldman (1983). Both models agree on the importance of encoding, storage, retrieval, and integration processes and the factors that influence them. For instance, DeNisi et al. pro­ posed that the purpose of appraisal will influence information search (see also Feldman & Lynch, 1988), and that this will have consequences for the later ac­ cessibility of this information. Feldman proposed that performance-relevant and performance-extraneous categorizations, as influenced by the relative accessibil­ ity of knowledge structures, will be used in either an automatic or controlled process that is involved in encoding information and making on-line judgments. These categorizations have consequences for later judgments, because they are differentially diagnostic for responses to different scales. The two models differ in emphasis and in the situations they consider. Each might be elaborated in the direction of the other.

ACCURACY IN PERFORMANCE APPRAISAL

It probably is not fair to claim that social cognition only recently discovered the issue of accuracy. Funder (1987), Kruglanski (1989), and Jussim (1991) made it clear that accuracy in interpersonal perception has been of concern to social psychologists for many years. Certainly, continuing interest in stereotyping processes and their avoidance (Devine, 1989; Fiske & Neuberg, 1990) has evinced a concern with accuracy in addition to social justice, as has research on the for­ mation of elaborated, individuated person representations. The common method­ ology of social cognition studies of person perception, showing that person representations can be influenced by momentarily accessible constructs (Bargh, 1984, 1989; Srull & Wyer, 1989; Wyer & Srull, 1980, 1986) and that (like stereo­

35 2

FELDMAN

types) spontaneously formed impressions may not be corrected for attributional/logical “errors” unless special conditions exist (Gilbert, 1989), testifies to an implicit concern with the potential lack of accuracy in daily life. However, it is fair to say that major attention has focused on descriptions of the processes of social perception and cognition. Tests of these processes typically involved the production of systematic inferential errors (e.g., Taylor & Fiske’s [1978] semi­ nal “token” effects, or Bargh & Pietromonaco’s [1982] demonstration of sublimi­ nal priming). These studies and others like them have given rise to widespread belief in the inaccuracy of social perception, as has similar research in the area of decision theory (Christansen-Szalanski & Beach, 1984; Jussim, 1991). In fact, the typical social cognition direct study can not provide evidence for accuracy, because the stimuli used are necessarily ambiguous. Recently, however, interest in the direct study of accuracy has reemerged (Kruglanski, 1989). In contrast, studies of performance appraisal in organizational psychology typi­ cally have been concerned first and foremost with accuracy. This is obviously because performance ratings are used for decisionmaking, with real impact on the welfare of individuals and organizations. As noted, early research focused on rating criteria such as “halo error,” simply because no accuracy criterion was available to organizational researchers. Rating-based measures were assumed to reflect error, and efforts to construct improved measurement formats were aimed at reducing it. Cognitive models of the rating process tended to focus on errors and biases, simply because these were traditional concerns of both fields. However, a conceptual advance took place when these cognitive models were combined with Borman’s (1978) demonstration that it was possible to study accuracy in the laboratory, albeit under a limited set of circumstances. The study of conditions promoting more or less accurate performance judgments became central to ap­ praisal research (DeNisi & Williams, 1988). Definitional Issues: Conceptual and Operational

Kruglanski (1989) presented three definitions of accuracy: correspondence with a criterion, consensus among judges, and adaptive valve in a given environment. Funder (1987) did not discuss accuracy in social perception directly. Rather, he concentrated on showing how the “illusions” and “errors” portrayed in experimental settings may reflect processes having adaptive value in the real world. The im­ plication was that substantial accuracy must exist for adaptation to occur at all. Likewise, Jussim (1991) did not deal with the definition of accuracy directly, but his discussion implied acceptance of the “correspondence with a criterion” model, the criterion in this case being the target’s actual behavior or attributes (whether preexisting or determined by the social perception process via selffulfilling or self-maintaining prophecy effects). In contrast, organizational psychologists have paid a great deal of attention

7.

PERFORMANCE APPRAISAL

3 53

to the conceptual and operational definition of accuracy. The basic paradigm is to present information about performance in the form of verbal scenarios or video­ taped performance episodes. These stimuli are constructed specifically to present evidence of high- or low-quality performance on one or more dimensions. Single or multiple targets may be employed. Consequently, both the occurrence (or nonoccurrence) of specific behaviors and the evaluation of performance (both overall and on specific dimensions) can be assessed. Often, the criterion for ac­ curacy is based on “expert” ratings of target person performance, as well as either recognition or recall of specific behaviors. The use of expert judgments, although not without controversy, allows a number of different indices of judgment ac­ curacy to be computed (Sulsky & Balzer, 1988): 1. D2, the squared difference between ratings and true scores, averaged over ratees and rating dimensions. This index can be decomposed into four independ­ ent sources of variance, as outlined by Cronbach (1955). These variance compo­ nents often are used separately, because they potentially relate to different sources of accuracy at a process level: a. Elevation, the differential grand mean. b. Differential elevation, the differential main effect of ratees (across dimen­ sions). c. Stereotype accuracy, the differential main effect of dimensions (across ratees). d. Differential accuracy, the differential interaction of targets x dimensions. Becker and Cardy (1986) showed that each term has both variance (i.e., differ­ ence score) and correlational components, and that these are independent. Thus, accuracy depends both on being able to produce a point estimate of a ratee’s stand­ ing on a dimension and being able to rank order ratees along that dimension. These accuracy components may have different sources. 2. Distance accuracy, the average absolute deviation between ratings and true scores averaged over dimension and ratees. 3. Borman’s (correlational) differential accuracy, based on the Z-transformed correlations between ratings and true scores, averaged across dimensions. Sulsky and Balzer (1988) pointed out that the purely correlational measures cannot be said to reflect accuracy, because they do not provide information as to the level of discrepancy between true scores and ratings, important for deci­ sionmaking in many contexts. Rather, they provide information as to rater valid­ ity, which is a necessary, but not sufficient, condition for accuracy. Along with accuracy in judgment, memory accuracy is important for perform­ ance appraisal. Besides the once common assumption that ratings depend on memory for behaviors at the time of rating, a number of appraisal instruments

354

FELDMAN

(e.g., BOS, BARS, BES, forced-choice) are thought to either tap recognition memory directly or depend on recognition in a straightforward way. Lord (1985b) proposed the use of two measures of recognition accuracy based on signal detec­ tion theory. Sensitivity, or d ' , is essentially the degree to which raters can dis­ tinguish between behaviors they have observed and behaviors belonging to the same category but have not been observed. A second measure, B, is the rater’s decision criterion (threshold) for choosing a response. Both are related to the “be­ havioral” accuracy of ratings, but in different ways. For example, d' may reflect rater categorization processes, whereas B may be sensitive to instructional manipu­ lations or rater motivation. As this discussion suggests, different accuracy criteria may be reflective of different processes, all of which contribute to overall accuracy in real-world set­ tings. For instance, Jussim (1991) discussed the ways in which social beliefs may relate to judgments of a target person’s behavior or attributes. For example, if information about a target leads to the target’s categorization, and that category has predictive validity for actual behavior or other stimulus attributes, both a high hit rate and a high false alarm rate will result, but accuracy will be higher than if a different (or no) categorization were made. A similar argument may be made for rating data. Lord and his co-workers demonstrated that accuracy in the per­ ception of a leader’s behavior was mediated completely by a “general leadership impression,” which created both a high hit rate and a high false alarm rate. Obvi­ ously, false alarms are undesirable for certain appraisal purposes—feedback, for instance, or documentation; but the findings do show that both accuracy and some sorts of bias can emerge from a single process. These measures reflect a variety of accuracy concepts, all of which are com­ patible with either an “agreement with criterion” or “consensus” definition of ac­ curacy. More precisely, they represent different concepts as to the source of accuracy. For example, Nathan and Alexander (1985) discussed the role of in­ ferential accuracy, the degree to which one can infer behaviors or performance levels not actually observed, based on the validity of one’s implicit theory of the job and the validity of one’s categorization of the target person. In Nathan and Alexander’s conception, accurate and thorough observation of behavior is a neces­ sary, but not sufficient, precondition for accuracy. One individual simply cannot observe all the relevant behaviors of another, nor recall all of those that are ob­ served to generate either recognition responses or evaluative judgments. Mur­ phy (1991) likewise pointed out that both behavioral accuracy (as discussed by Lord) and classification accuracy are necessary. Jussim (1991) made clear that these criteria are not independent; accurate recognition may be a function of either direct recall or inference, which depends on categorization. Accurate prediction of behavior depends likewise on both observation and information processing, via social beliefs and implicit theories (Zalesny & Highhouse, 1992). It also should be noted that classification accuracy basically refers to a consensus definition of accuracy, because categorizations ultimately are based on consensual agreement

7.

PERFORMANCE APPRAISAL

355

about category systems and the nature of job performance. To the extent that accuracy of agreement with a criterion depends on classification, the agreement with criterion and consensus definitions are not independent. Feldman (1986b, 1992a) made the explicit statement that, for many jobs, appraisal is inherently subjective, based on value systems and concepts differentially shared throughout the organization. Without a commonly held, context-independent value and category system, one can expect neither consensus nor that component of be­ havioral accuracy/criterion agreement that is due to inferential processes. Sulsky and Balzer (1988) made essentially the same point in their discussion of the need to validate “expert judgments” before accepting them as criterion standards. (Ex­ pert ratings, as it happens, seem to have substantial validity; Smither et al., 1989.) The preceding discussion raises two interesting points: First, that the relation­ ship between rating errors produced by cognitive processes operating on obser­ vations and various accuracy scores may be positive rather than negative. For instance, a high hit rate for recognition-based measures or accurate category-based evaluations may be accompanied by substantial false alarm rates or halo errors, respectively. Second, the utility of ratings, despite these errors, may be substan­ tial. Utility must be demonstrated by construct validity procedures, rather than an exclusive focus on accuracy. Consideration of the process sources of accura­ cy and specific kinds of inaccuracy (e.g., of classification, inference, or memory) seems to be the key to improving ratings.

Rating "Errors" and Accuracy

Murphy and Balzer (1989) and Becker and Cardy (1986) conducted studies of the relationship between errors (e.g., halo, leniency) and accuracy. Murphy and Blazer’s meta-analysis investigated the correlation between Cronbach’s four ac­ curacy components and two measures of each of three types of error: halo, lenien­ cy, and range restriction. Relationships were largely trivial, except between halo and differential accuracy, where increasing halo was associated with greater ac­ curacy. The authors pointed out that halo, whether measured by dimensional in­ tercorrelation or variance of ratings across dimensions (within-rater, across targets), does not necessarily reflect error, because there is a true halo level reflect­ ing the actual intercorrelation of performance dimensions. Fisicaro (1988) directly examined this issue, finding that ratings displaying both more and less intercorre­ lation than the “true” score relations are less accurate; the positive linear rela­ tions observed earlier (Cooper, 1981) are due to the use of a linear correlation coefficient to model a quadratic relation with a nonzero, probably moderate, in­ flection point. Becker and Cardy pointed out that alternative measures of both halo error and accuracy lead to positive, negative, or no relations, depending on whether variance-based and/or correlational measures of each are applied.

356

FELDMAN

These reflect relative differences between ratees or the absolute deviation of rat­ ings from true scores, respectively. The relations between halo and accuracy measures are quite complex, but may be illuminated by considering both: (a) how person impressions are organized, and (b) how accuracy scores are computed. Considering the latter, it is crucial to remember that each accuracy component score is independent of every other, in both its variance and correlational terms. Considering the former, one can adopt provisionally Srull and Wyer’s (1989) concept of separate trait and evaluative impressions. The trait impressions are constructed based on the sample of ob­ served behaviors, because they are diagnostic of concepts accessible at the time the behaviors are encountered. Evaluative impressions are based on the implica­ tions of all behaviors, regardless of whether they are associated with a trait con­ cept. For the present purpose, I regard impressions of performance on multiple dimensions as equivalent to trait concepts. It is problematic, and an empirical question, whether performance dimension concepts are accessible to subjects at the time they encounter the stimulus information. When ratings are required, it may be assumed that subjects use the most diag­ nostic of the accessible information they possess to generate ratings. If they have formed judgments on performance dimensions, these will be accessed for rat­ ings. However, these impressions will not necessarily be independent. If, in the subject’s implicit theory, the dimensions are associated, observed behaviors con­ tributing to one will also contribute to another, and the resulting ratings will be correlated. The more the subject’s assumed correlation matches the actual corre­ lation of the behaviors, the more accurate will be the resulting ratings (in the sense of differential accuracy), because variability in the ratings will match more closely variability in the actual occurrence of the behaviors. However, what if no specific performance dimension representations are formed? The rater most likely will use the evaluative impression, which is the most accessible and is at least moderately diagnostic for the rating judgments (Feld­ man & Lynch, 1988; Srull & Wyer, 1989). To the extent that task/job behaviors have influenced this overall impression (as opposed to extraneous characteristics or behaviors), valid information will be contained in it. Suppose this global evalu­ ation is used to construct dimensional ratings in accord with the rater’s implicit theory of performance; to the extent that this theory is valid (Jussim, 1991), rat­ ings on the separate dimensions will vary (and covary) as if they were based on actual observation. Again, differential accuracy would be relatively high, and halo would be empirically correlated with it, as in Fisicaro’s (1988) data. The argument is also consistent with the data presented by Murphy and Balzer (1989) and Becker and Cardy (1986). Why, then, would halo measures not correlate with differential elevation? The answer is that they might. Although Murphy and Balzer (1989, Table 3) found the average correlation to be low, Becker and Cardy (1986) found correlational measures of halo to relate positively to the correlational component of differential

7.

PERFORMANCE APPRAISAL

357

elevation (i.e., higher interdimensional correlations were associated with more accurate overall rankings). It may be, as Becker and Cardy suggested, that rela­ tively large between-ratee differences in overall performance contribute to sub­ stantial differences in global evaluations. If such were the case, and if global evaluations of each ratee were accessed to construct each rating, one would ex­ pect both halo error and accurate differential elevation, because between-ratee variance would be high. At the same time, if overall evaluation and an accurate implicit theory of performance were used to generate ratings, one would expect correlations between variance-based halo measures and differential accuracy. Be­ cause differential accuracy is independent of differential elevation, biases produced by the use of an overall evaluative impression would not affect this criterion as they might affect differential evelation. Because variance-based halo measures generated by an accurate implicit theory reflect actual average differences in per­ formance levels across dimensions (and within ratees), theory-based or top-down dimensional ratings constructed from an overall impression would be relatively accurate, as are predictions made by any other valid theory on the basis of given input parameters. This is what Becker and Cardy’s (1986) data showed; the meta­ analysis of Murphy and Balzer (Table 3, row 1) is at least consistent with the argument. Before I conclude that halo in ratings indicates at least some degree of veridi­ cal observation and judgment, however, another issue must be considered. Jus­ sim (1986, 1991) discussed self-fulfilling and self-sustaining prophecies, which under some conditions induce behavior confirming initial expectations. Feldman (1986a) argued that an initial impression, perhaps based on valid information but perhaps also based on stereotypes or other misleading beliefs, nevertheless may produce differential treatment of employees (Eden & Shani, 1982; Graen, 1976). Positive-expectation employees are given challenges, opportunities, support, and feedback that others are not. To the extent that this treatment persists, employees will have developed skills and abilities as a result, in accord with expectations and consistent with the nature of the job. Thus, an employee’s true performance will come to reflect what may have been (at least partially) systematic error in ratings. This is not a problem of performance appraisal judgments, but of expec­ tations and social behavior in the workplace. Construct Validity of Performance Appraisals

Direct accessment of accuracy in appraisal ratings is limited to laboratory studies. However, if ratings made in natural circumstances contain at least some truescore variance, one would expect to find certain patterns of relationships between ratings and other variables. For instance, ratings (and rankings, presumably based on overall evaluations) should show convergent validity with other criteria, such as quantity or quality of production. They also should be predictable by tests of ability in specific patterns appropriate to the job in question; for instance, an

358

FELDMAN

aircraft mechanic’s ratings should correlate with measures such as mechanical aptitude and job-specific knowledge. These patterns would support the inference that relevant job behavior was influenced by appropriate knowledge, and was reflected in the judgments made by supervisors. Nathan and Alexander’s (1988) meta-analysis showed that both ratings and rankings are predicted well by a vari­ ety of ability measures, as are work samples and quantity of production, although quality measures seem less predictable. Ratings and quantity criteria, in fact, gave much the same information. Both Schmidt, Hunter, and Outerbridge (1986) and Borman et al. (1991) found ratings to be determined by job knowledge, task profi­ ciency, and other sources of interindividual variance. In addition, MacKenzie, et al. (1991) found that “extra-role” behavior (activities that are not part of a for­ mal job description, but which add to the effective functioning and positive at­ mosphere of a workplace) add variance to appraisal ratings over and above objective performance indices. Together, these studies provide evidence that rat­ ings contain valid performance variance and provide support for the contentions of accuracy by Jussim (1991), Kruglanski (1989), and Funder (1987). In subse­ quent sections, I examine the social-cognitive aspects of the appraisal process, and factors that influence the degree of accuracy one may expect from ratings. HALO AND BIAS IN PERFORMANCE APPRAISAL

The problems of “halo error” and bias due to information extraneous to perform­ ance (typically, membership in a stereotyped category) usually are not consid­ ered together. It seems useful to adopt such a strategy, however, because it is likely that both phenomena stem from a common source: observed behavior and other characteristics of the ratee contributing to a person impression. The con­ struction of this impression is governed by implicit theories of both persons and job performance, theories containing both conceptual and affective information. The resulting impressions and organized memories are used (according to the principles of diagnosticity and accessibility discussed earlier) to generate ratings of varying degrees of accuracy. Halo

As discussed by Balzer and Sulsky (1992), there is no single, generally agreedon conceptual definition of halo error. Two distinct concepts traditionally have existed: 1. Judgments along a number of ostensibly distinct dimensions are based on an overall evaluative impression. 2. Judgments along a number of ostensibly distinct dimensions are based on beliefs about the similarity between, or implicative relationships among, those dimensions.

7.

PERFORMANCE APPRAISAL

359

The former, called general impression halo by Balzer and Sulsky, may be said to result from the retrieval of an overall evaluative impression to generate rat­ ings. The latter, termed dimensional similarity halo, may result from either en­ coding processes (in which a given behavior or attribute may contribute to two or more trait/performance dimension impressions), or constructive inference processes, in which a given trait or performance dimension impression is used to infer a ratee’s standing on a dimension for which there is either no information or ambiguous information. Lack of conceptual agreement is reflected in a lack of agreement at the opera­ tional level. Both variance-based and correlation-based measures of halo exist (Balzer & Sulsky, 1991; Becker & Cardy, 1986; Murphy & Balzer, 1989). Both measures reflect limited implications of dimensional similarity and general im­ pression halo; that is, there is less variance between dimensions and higher inter­ correlation across dimensions than would be expected on the basis of the true performance scores. Importantly, both variance and correlational indices should be computed within raters, because halo is defined as a phenomenon at the in­ dividual rater level (Balzer & Sulsky, 1992). But neither index alone can distin­ guish between halo that is caused by a general impression driving ratings and halo that is caused by systematic distortion (Schweder, 1977; Schweder & D’An­ drade, 1979) due to either conceptual similarity among dimensions or implicit theories of performance. It also should be noted that simple variance or correla­ tional indices cannot, in and of themselves, indicate the degree to which halo equals halo error. To do that, an estimate of true halo (Cooper, 1981), the actual variance and covariance among performance dimensions, is necessary. Overall Impression Halo

Evidence that halo can originate in an overall impression of the ratee was provid­ ed by Nathan and Lord (1983). They found that a general impression of the ratee contributed significant variance to dimensional ratings, even though performance on three of five dimensions was in a direction opposite to the ratee’s overall per­ formance. Although performance on each specific dimension was perceived ac­ curately, it also was true that an overall favorable or unfavorable impression at least partially mediated the rating on each separate dimension. There was also an effect of overall impression on recognition accuracy, such that subjects were more likely to falsely recognize behaviors consistent with this general impres­ sion. Cardy and Dobbins (1986) provided evidence that extraneous variance (in this case, personal liking) may be incorporated into performance impressions. They compared performance information in which adjectives of differential favorability were varied orthogonally with performance to cases in which information about favorability was constant across performance levels. Orthogonal variation in favorability, although irrelevant to the ratings, significantly reduced differen­ tial accuracy.

360

FELDMAN

These results suggest that, at least for inexpert raters, liking is integral to per­ formance judgments. Other dimensions (discussed later under Bias) likewise may be difficult to separate from performance. To the extent that information along these dimensions influences evaluative impressions and those impressions are ac­ cessed to construct ratings, extraneous factors will add to illusory halo. Tsui and Barry (1986) showed that liking or disliking is a particularly potent source of rating variance in field settings. Thus, it is likely that in the construction of an overall evaluative impression, both performance-relevant and performanceirrelevant information will be considered. The relative contribution of these types of information and the degree to which an overall evaluative impression is ac­ cessed and integrated into dimensional and overall ratings will determine the degree of overall evaluation halo and its relation to various components of accuracy.

Dimensional Similarity Halo

The second source of halo, dimensional similarity, is more complex. I include in this category halo based on implicative relationships among performance dimen­ sions, as well as assumed similarity. For example, planning and organization may be seen as requiring similar behaviors, and may be regarded as similar aspects of performance. In contrast, concern for quality and promptness may be perceived as related via a common high-order dimension such as conscientiousness. Thus, promptness might imply quality, even though the two are not at all similar. Like­ wise, a person may be rated similarly on apparently dissimilar aspects of per­ formance, because both are part of a group stereotype. For instance, job knowledge and motivation might be rated similarly, because trait concepts such as intelli­ gence and ambition (or their opposites) are part of a stereotype fitting the em­ ployee. Such covariance might be produced by either biased encoding of information or biased retrieval, as well as by category-based inferential process­ es (Darley & Gross, 1983). The role of implicit theories, subject-matter expertise, and knowledge of in­ dividual target persons was illustrated by Kozlowski, Kirsch, and Chao (1986) and Kozlowski and Kirsch (1987), who had subjects varying in baseball knowledge rate several aspects of the performance of high- and low-familiarity players. They also assessed conceptual similarity among performance dimensions, rating scale intercorrelations, and true-score intercorrelations. If one assumes that knowledge of individual players is highest in the highknowledge/familiar player cell, lowest in the low-knowledge/unfamiliar player cell, and intermediate in others, their results (in broad outline) reveal that im­ plicit theories (or conceptual similarity) are used to generate reasonably accurate ratings when rater knowledge is moderate (i.e., low-knowledge/familiar player and high-knowledge/unfamiliar player conditions), and that rating accuracy de­ pends on the accuracy of the implicit theory. When knowledge is high and the

7.

PERFORMANCE APPRAISAL

361

player is familiar, individuated impressions generate ratings with little halo and high accuracy. High knowledge coupled with an unfamiliar player produces more halo and some accuracy, as does the low-knowledge/familiar target combination. However, low knowledge and unfamiliar targets lead to ratings that are nearly random. The inference is that even theory-based heuristic reasoning needs some input on which to operate. In the latter case, the necessary knowledge is simply unavailable. The process seemed well handled by Jussim’s (1991) reflectionconstruction model, which postulated that accuracy may stem either from accurate perceptions of target attributes (individuated impressions) or a valid theory of social behavior (in this case, task performance). Kozlowski and Mongillo (1992) demonstrated that conceptual similarity sche­ mata and implicit theories can simultaneously show substantial between-person differences in content and substantial accuracy. Their results agreed with studies of event categories (e.g., Billman & Knutson, 1990), in that schemata are shown to be highly responsive to experience and to reflect patterns of association in the stimulus environment. Thus, an important source of both agreement between ob­ servers and accurate judgment is the degree to which implicit theories are based on common experience and have similar contents (Jussim, 1991; Kenny, 1991). This also underscores the role of subject-matter expertise in promoting accurate and unbiased performance appraisals. Bias

In contrast to the debate surrounding definitions of halo, researchers seem to have agreed on the definition of bias: differences in the judgments of persons (or their performance or behavior) associated with membership in a specific socially de­ fined category. Operationally, this straightforward definition leads to equally straightforward experimental designs. Subjects rate stimulus persons whose be­ havior or performance is identical, but who differ in category membership. Any differences in appraisals are attributed to the effects of category membership. More complex designs vary other attributes as well (e.g., leadership style or gender role identification of tasks: Eagly et al., 1992; Swim et al., 1989), but all are based on the idea that bias is evidenced by differences in rating among otherwise identical individuals. In this, bias research has echoed early work on stereotyp­ ing and performance appraisal as influenced by attributional processes (Feldman, 1972; Feldman & Hilterman, 1975, 1977). The use of category membership as input to a performance judgment is considered a bias by virtue of the explicit demand to evaluate only the behavior, performance, or product immediately present. To the extent that the subject is asked to infer to unseen behaviors or attributes for which category membership is diagnostic, the lack of attention to the target’s category (indexing base rates for the attribute or behavior in ques­ tion) may be considered a bias (Locksley, Hepburn, & Ortiz, 1982; Tversky & Kahneman, 1982).

362

FELDMAN

Typically, attention is focused on underevaluation of minority groups (e.g., Black, female, the aged), echoing the traditional concerns of social and organiza­ tional psychology. However, at a process level, one question raised by this oper­ ational definition of bias is whether “majority” group members are overevaluated or minorities underevaluated. A related question, raised in prior stereotype research but typically not addressed in either social or organizational studies, con­ cerns conditions in which minority members are overevaluated (Feldman, 1972; Feldman & Hilterman, 1975). A second set of questions concerns the degree to which the bias discovered in laboratory studies is indicative of bias in the real world (e.g., Wendlken & Inn, 1981). Given that conditions in “natural” rating situations involve extended contact, outcome dependence, relatively higher levels of task knowledge, and so on, there is reason to believe that judgments in those conditions will be made on a more individuated, thoughtful basis (Ilgen & Favero, 1985; Murphy, Herr, Lockhart, & Maguire, 1986). However, the issue is complicated by the fact that, although ratings in the laboratory might be taken as reasonably veridical indica­ tions of judgments, ratings under real-world circumstances are influenced by many factors other than the stimulus person’s attributes, behavior, or performance (Longnecker et al., 1987). Another complicating factor is that ratings in real organiza­ tions cannot be assumed to be based on equivalent information; actual minority/majority differences influence ratings, and must be taken into account. Finally, one must consider the causes of any rating effects. As in the case of halo error, rating bias may stem from either or both of two sources. The first is the incorporation of affective responses toward the category of persons into rating construction. This bias may stem from category-based affect (Fiske & Pavelchak, 1986) that is incorporated either directly into the performance judgment or into the person impression. Second, implicit theories (or stereotypes) may bias inferences either directly (e.g., some person of Group Y is more/less likely to be intelligent than some person of Group X) or via biased encoding of behavior (Darley & Gross, 1983). Attributional processing at this point may lead to over­ evaluation of minority persons, as previously noted, as well as to underevaluation. Gender

In recent years, the issue of gender role stereotypes has become increasingly im­ portant in matters of organizational and social policy. Laboratory research has assumed a great deal of importance in resolving charges of gender bias (Fiske, Bersoff, Borgida, Deaux, & Heilman, 1991). However, there is little evidence that simple, direct effects of gender on evaluations occur alone. Rather, when a gender effect does occur, it seems to be the result of relatively complex attribu­ tional processes. Swim et al. (1989) found that, contrary to popular belief, the simple effects of gender differences on performance evaluations were negligible. These results

7.

PERFORMANCE APPRAISAL

363

echo those of field studies of appraisal in real jobs (Pulakos, White, Oppler, & Borman, 1989; Thompson & Thompson, 1985; Wexley & Pulakos, 1982), which found no male/female differences in ratings. This is reasonable, because females as a social category are not likely to be disliked (as are, for instance, ethnic minori­ ties). The simple incorporation of overall affect into ratings would be unlikely to produce anti-female bias. In fact, there has been evidence that, in some cir­ cumstances, women receive more favorable judgments than men (e.g., Bemmels, 1988; Schneer & Reitman, 1990). However, there has been evidence for the impact of gender role stereotypes and social norms on judgments of female employees (Deaux, 1984), although this was moderated heavily by the presence of individuating information and the nature of the comparison requested. Sackett, Dubois, and Noe (1991) found in a field study that women were rated lower as their percentage in a work group decreased. However, this finding was not replicated for Blacks. Although super­ ficially this finding seems to reflect Taylor and Fiske’s (1978) studies of tokenism and the causal impact of novelty on perceptions, several factors should be con­ sidered: First, stereotypes do not necessarily reduce evaluations. Not only do positive stereotypes exist, but even in the case of negative stereotypes, unexpect­ edly good performance can lead to overevaluation (Feldman, 1972; Feldman & Hilterman, 1975). Second, high-performing groups with one or a few female mem­ bers should attribute more responsibility for the favorable outcome to the novel members, according to the Taylor and Fiske (1978) arguments. Third, external factors may bias ratings. The lower ratings of novel females may be an artifact produced by deliberate overevaluation where women exist in larger numbers. It is also possible that ingroup/outgroup downward biasing factors (e.g., the “minimal group effect”) produces lower ratings of women in organizations where they are less frequent (and perhaps more recent) members. Likewise, Olian, Schwab, and Haberfield (1988) found a small pro-male bias in hiring recommendations. Male/female differences accounted for only 4% of the variance overall, however, whereas qualification manipulations accounted for 35%. Also, effect sizes were more pronounced in within-subjects than betweensubjects designs, suggesting that making contrasts highly salient results in great­ er differentiation among stimulus persons on whatever grounds are available. Such contrast effects are likely whenever the study format implies that comparisons between stimulus persons are the focus of concern (Schwarz & Bless, 1991). The foregoing should not be taken to imply that women are never disadvan­ taged in evaluations. For instance, Eagly et al. (1992) showed that female lead­ ers were rated worse than their male counterparts when they exhibited autocratic or directive behaviors, or when they occupied male-dominated roles and were evaluated by males. This could be explained parsimoniously as a stereotype-based contrast effect; if women are expected to be cooperative and participative, au­ tocratic behavior (or the occupation of positions implying autocratic behavior) leads to their being seen as highly autocratic, in accord with Jones and Davis’

364

FELDMAN

(1965) “principle of augmentation” (see also Feldman, 1972; Feldman & Hilterman, 1975). Heilman, Martell, and Simon (1988) showed that these processes can produce either over, under, or equal evaluation of males and females. Also in keeping with the idea that women’s evaluations are mediated largely by stereotyping processes are the results of Heilman and Martell (1986). Specifi­ cally, decisions about male and female applicants for a traditionally male job ex­ hibited a pro-male bias unless subjects had been exposed to highly job specific counterstereotypic information. Dobbins, Cardy, and Truxillo (1988) found that subjects holding traditional female stereotypes evaluated female ratees lower and less accurately (in terms of differential elevation) than did those with nontraditional beliefs. This anti-female bias occurred in ratings of actual as well as hypothetical stimulus persons. However, it was only evident when subjects be­ lieved their ratings were made for administrative, rather than research purposes. Dobbins et al. speculated that the administrative purpose required inferences about future performance, rather than simply evaluations of behavior, making gender stereotype information perhaps diagnostic for the judgment. This is consistent with the results of Glick, Zion, and Nelson (1988), who found that gender dis­ crimination in hiring decisions was mediated in part by beliefs about individual characteristics (as influenced by both stereotypes and individuating information), and in part by gender-related implicit theories of the jobs. The role of gender and occupational stereotypes in conjunction with individuat­ ing information and attitudes toward women in occupations was explored by Maur­ er and Taylor (in press) in a field study of student’s ratings of professors, extending the Dobbins et al. (1988) earlier research. Perceived masculinity/femininity of a sample of male and female professors was measured, as were causal attributions for good and poor performance, attitudes toward women as instructors, and mascu­ line/feminine aspects of an “effective college instructor” category. Also meas­ ured were the degree to which each instructor exhibited each of a set of 30 positive and negative behaviors, the response latency for each behavior (corrected for read­ ing time), and evaluative ratings on nine performance dimensions. Neither the masculine/feminine job requirements nor the trait ratings assessing masculini­ ty/femininity formed a unidimensional scale. This suggests that concepts of gender are categorical, rather than continuous, and at least partially independent. Instructor gender, in and of itself, was unrelated either to appraisals or to any of the possible mediating variables. However, evaluations of female instructors were related positively to perceptions of their masculinity. In contrast, evalua­ tions of male instructors were correlated positively with perceived masculinity only when subjects had negative attitudes toward women. Among subjects with positive attitudes toward female instructors, the relationship was negative. Ap­ parently, the affective value of one’s nonteaching behavior influences both evalu­ ations and recognition/frequency responses, and the direction of affect generated depends both on who performs the behavior and the perceiver’s value system. The conclusion that categorization as a generally effective or ineffective in­

7.

PERFORMANCE APPRAISAL

365

structor mediated evaluations is supported by the relation between frequency es­ timates pertaining to each behavior and response times. The time taken to judge positive behaviors was less when the behaviors were more frequent, whereas the time to judge negative behaviors was less when the behaviors were less frequent. Direct ratings of performance quality showed similar relationships with response time. Interestingly, the accessibility of female’s positive behavior (as measured by response time) increased with perceptions of their masculinity, whereas the accessibility of male’s positive behaviors did not. Raters with positive attitudes toward female faculty likewise showed greater accessibility of positive behaviors for females, whereas those with negative attitudes showed less. Accessibility of positive behaviors for male ratees was intermediate and unaffected by either per­ ceived masculinity or attitude. These findings indicate that performance appraisals can be influenced by both affect and schematic inferences, and that schemata for male and female incum­ bents may differ in accord with more general value or ideological systems. The results for causal attribution are in accord with this conclusion. It seems, then, that gender-stereotyped decisions may be overcome by individu­ ating information and/or counterstereotypic persuasive communications. However, in all of these studies, subjects had both the motivation and capacity to process in­ formation (Devine, 1989), and the individuating information was processed in the context of the (probably) automatically elicited stereotype (Bargh, 1989; Devine, 1989). It is likely that a number of factors influence the outcome of interactions be­ tween men and women in the workplace and the role of stereotyping and affec­ tive processes in those interactions. Individuals who are highly gender-schematic (Bern, 1981) will be likely to process information in terms of gender role expec­ tations, and thus show more extreme stereotype confirmation/contrast effects (Fong & Markus, 1982; Markus, 1977). Those who are most strongly schematic also might have a stronger affective response to women whose behavior contradicts traditional expectations (Pazy, 1986). Initial evaluations may be a small, but crit­ ical, part of a process in which women may have a more difficult time establish­ ing a mentored relationship (e.g., as in Graen’s vertical-dyad model, 1976) and accumulating power-producing resources (Ragins & Sundstrom, 1989). Affirm­ ative action, designed to overcome these difficulties, easily can create outgroup perceptions and may, in fact, impact negatively on women’s job- and self-concepts unless specific information to the contrary is presented (Heilman & Herlihy, 1984; Heilman, Lucas, & Kaplow, 1990). The result might be self-fulfilling prophe­ cies of the sort described by Jussim (1986, 1991) and Ilgen and Youtz (1986). Race

In contrast to research on gender effects, laboratory studies of racial bias in per­ formance appraisals have been relatively infrequent since the 1970s (Kraiger & Ford, 1985). However, studies of real-world rating bias have been increasingly

366

FELDMAN

frequent. These typically are aimed at assessing the existence of bias, rather than at questions of process. Kraiger and Ford’s (1985) meta-analysis revealed that Black raters tended to rate Blacks higher than Whites, whereas Whites tended to rate Whites higher, consistent with both a stereotype and an affective bias model. They also found greater effect sizes in field than laboratory settings, suggesting that either norms of social desirability or equalitarian values influenced either overt ratings or judg­ ments in the largely student laboratory samples. Ford, Kraiger, and Schectman (1986) found that, although objective performance differences occurred as a func­ tion of race (Blacks performing more poorly than Whites), cognitive measures had the highest effect sizes, rating-scale measures the second highest, and objec­ tive performance indices the lowest. This pattern of results is consistent with two factors: regression toward the mean in predicting performance from cognitive ability measures, and exaggeration of real differences via perceptual phenome­ na, a result of fundamental category-formation processes (Hamilton, 1979; LeVine & Campbell, 1972). These results may suggest qualifications to the arguments of Wendelken and Inn (1981), Ilgen and Favero (1985), and Murphy, Gannett, Herr, and Chen (1986), who claimed that highly controlled laboratory studies and attentionfocusing “paper people” designs lead to overestimates of social cognition-based biases. However, a number of confounding factors may have contributed to Kraiger and Ford’s results. First, Black and White ratee samples may not have been equivalent in education or experience, producing real performance differ­ ences. Second, Black and White raters typically do not rate the same samples. Thus, if people were assigned differentially to Black and White supervisors (es­ pecially by race), it may have been that either rating standards came to differ or actual ability differences existed. As in the case of gender bias, simple race effects were small. Waldman and Avolio (1991) found only 3% of the variance in appraisals was due to race, and when ability and experience of the ratee were taken into account, this effect disappeared. Pulakos et al. (1989) examined rat­ ings of Blacks and Whites (as well as males and females) in a military sample using Black, White, male, and female raters for the same ratees. Differences were generally small. However, Whites received higher ratings from both Black and White raters, although Blacks were rated more highly by Black than by White raters. Also, the Black-White rating difference was reversed on one of three per­ formance dimensions; whereas Whites were rated more highly than Blacks on “technical skill” and “personal discipline” dimensions, Blacks received higher rat­ ings than Whites on “military bearing.” The fact that this pattern was consistent across Black and White raters argues against a stereotype-contrast explanation. Whether anti-Black, pro-White, anti-White, or pro-Black rating biases also exist is still in question, however. Sackett and Dubois (1991) extended the Pulakos et al. (1989) study to include a civilian sample, with largely the same results for overall performance. Furthermore, their reanalysis of the Kraiger and Ford (1985)

7.

PERFORMANCE APPRAISAL

367

and Pulakos et al. (1989) data revealed that the differences in rating patterns asso­ ciated with Black raters occurred primarily in laboratory studies using undergraduate raters and peer raters (Kraiger & Ford) and peer raters (Pulakos et al.). This sug­ gests that the expertise associated with supervisory positions can mitigate against bias, presumably by creating common standards of judgment and chronically ac­ cessible categories of performance and behavior. This conclusion was supported by the extensive analyses in Oppler, Campbell, Pulakos, and Borman (1992). Other Biases

Besides race and gender, other socially defined categories are associated with stereotypes relevant to performance appraisal. Of these, age and physical attrac­ tiveness are among the most intensively researched. As before, two sources of bias may operate. That is, affect associated with category membership may be incorporated directly or indirectly into evaluations, or stereotypic influences may bias the encoding of behavior and subsequent inferences, contributing to judg­ ments on ambiguous or unobserved performance dimensions. Likewise, individu­ ating information may operate in concert with the stereotypic to produce either stereotype confirmation or contrast effects, resulting in over- or underevaluation of an individual’s performance via attributional reasoning. Eagly, Ashmore, Makhijani, and Longo’s (1991) review of the physical at­ tractiveness stereotype was consistent with these arguments. Attractive stimulus persons were thought to be higher in social competence, but also more likely to be vain. They were seen as better adjusted, more influential, and more in­ tellectually competent as well, although effect sizes were smaller. No differences were found on integrity or concern for others, although a substantial effect on overall evaluation was found. Factors influencing the strength or vividness of the attractiveness manipulation (e.g., the use of color photos) increased the ef­ fect, but individuating information also moderated people’s inferences. In short, physical attractiveness seems diagnostic for certain judgments, but not others. People are responsive to the information given, which moderates their use of the stereotype when more diagnostic information is available. Attractiveness seems to interact with information about gender in stereotypeconsistent ways. Heilman and Stopeck (1985a) found that attractive men described as successful in a corporate environment were seen as more able than the less attractive, whereas attributions of ability were reversed for attractive and unat­ tractive women. In a study of performance appraisals of managerial and nonmanagerial job holders, Heilman and Stopeck (1985b) found that attractiveness improved the evaluations of women in nonmanagerial positions, but reduced them for female managers. Male’s appraisals were unaffected. The mediating variable seemed to be perceived masculinity/femininity; femininity was enhanced by at­ tractiveness, whereas masculinity was not, and so the more attractive women may not have “fit” the traditionally masculine managerial position.

368

FELDMAN

Finally, evidence exists that involvement and issue importance may override the biasing effects of attractiveness. Downs and Lyons’ (1991) natural observa­ tion study of bail decisions revealed that more attractive defendants received lower bail and fines for misdemeanors, but not for felonies. The deeper processing in­ volved in making judgments for more serious crimes apparently may overcome whatever effects of attractiveness or stereotypical inferences operate under lowinvolvement circumstances. Age bias operates similarly. Rosen and Jerdee (1976a, 1976b) detailed the nega­ tive job-related stereotype of older workers (relatively passive, reserved, and in­ flexible), and showed that it influences judgments in ratings of hypothetical others. Cleveland and Landy (1983) demonstrated that fit to the stereotypical age require­ ments of jobs influences evaluations; younger workers and workers performing in a stereotypically “young” way are evaluated more highly in jobs typically held by younger workers; the difference is less, or slightly reversed, for jobs typical­ ly held by older workers (see also Lawrence, 1988). Similarly, Rosen, Jerdee, and Lunn (1981) found that retirement recommendations were stronger for older workers unless they also were described as high performers. Lee and Clemons (1985) reported that age bias was reduced when specific performance informa­ tion was provided, and when a rating rather than an older/younger choice response was made. Thus, salient individuating information seems as highly diagnostic for judgments of older and younger employees as it does for male/female and Black/White judgments. However, evidence exists that discrimination may persist in the real world. Waldman and Avolio (1986) found a negative relationship between employee age and supervisory ratings (at least for nonprofessionals), despite a positive rela­ tionship between age and actual productivity. It may be that age stereotypes produced biased encoding or retrieval under natural circumstances, leading to stereotype confirmation biases (Darley & Gross, 1983; Kunda, 1990). Summary

It seems very likely that the sources of halo and bias are identical at a process level. Both may be linked plausibly to either the direct or indirect incorporation of affect into performance judgments, and/or reasoning guided by implicit the­ ories of jobs, performance, personality, or social categories (i.e., stereotypes). Reasoning governed by attributional heuristics (e.g., augmentation/discounting) is included in this second category. The operation of both principles is consistent with the ideas of accessibility and diagnosticity proposed by Feldman and Lynch (1988), and with the concept of automatic and controlled processing (whether central or peripheral) discussed by Gilbert (1989), Devine (1989), and Bargh (1989), contingent on motivation and capacity. Appraisal research on bias might benefit from a more process-oriented ap­ proach. Appraisal accuracy could be improved substantially if training and in­

7.

PERFORMANCE APPRAISAL

3 69

strument development focused on ways to disentangle category-based affective responses from work-performance appraisals, or permitted the automatic encod­ ing of performance-relevant behaviors via a conceptual system that was independ­ ent of stereotypical beliefs. Accountability processes that foster reconstruction of stereotypically biased judgments also could be investigated, both in simula­ tions and in field experiments. This research would be of value to social cognition investigators as well, for reasons beyond the ecological validity of laboratory studies. Relatively little is known about either encoding and retrieval processes over substantial time peri­ ods, or about the influence of relatively intense affective responses characteristic of ethnic bias. It is feasible to collect recall data in field settings, and, in many cases, reaction-time data also may be obtained. These would allow tests of process­ ing hypotheses under different conditions than can be produced in the laboratory. However, the question of discrimination and bias in appraisals goes beyond the issues that typically are considered in social cognition research, whether the­ oretical or applied. As Dipboye (1985) cogently observed, appraisers are not pas­ sive observers. They interact with subordinates, are responsive to social pressures and organizational climates, and act toward subordinates in ways intended to pro­ tect their personal feelings and avoid discomfort. This behavior not only influences ratings, but affects the interpersonal environment, influencing impressions formed by supervisors and the self-judgments of employees. These may influence subse­ quent performance and behavior (Feldman, 1986a; Ilgen & Youtz, 1986; Jussim, 1986; Snyder, 1984). Furthermore, even effects controlling small amounts of judg­ ment variance may have major social significance (Martell, Lane, & Willis, 1992). Interpersonal behavior and its consequences, as mediated by social-cognitive processes, may be studied in natural environments, as well as in the laboratory. The strong possibility exists that, by doing so, one can improve theory and prac­ tice and increase social justice as well.

PERFORMANCE APPRAISAL CONTEXT

A much-studied issue in performance appraisal is the effect of the context in which ratings are made. Context in this case has multiple meanings, however. One form of context effect is evaluative: Does the quality of one or more observed instances of performance influence the perceived quality of other performances? Another is appraisal purpose: Do ratings and judgments change when appraisals are made for different reasons? A third is the organizational context: What are effects of the contingencies, both administrative and interpersonal, under which judgments and ratings are made? This contextual factor includes accountability for ratings, as well as other possible motives of the rater. As is seen, the effects of these various appraisal-context issues may be under­ stood in terms of more fundamental impression formation processes. These

370

FELDMAN

processes include category construction and accessibility effects as influenced by appraisal purpose, and depth and direction of processing as influenced by moti­ vation. These have systematic effects on ratings (both level and accuracy) as well as memory organization and memory-judgment relationships. Evaluative Context

In real-world circumstances, part of the context for each rater/ratee pair is the ratee’s own performance history. It is often the case that a given performance by a ratee must be judged on its own merits, but with knowledge of prior per­ formance. Likewise, performance episodes may be judged retrospectively, in light of subsequent performance. Murphy, Balzer, Lockhart, and Eisenman (1985) and Murphy et al. (1986) investigated both cases. In the former case, contrast effects seemed to predominate. That is, average performance was rated worse when it was preceded by good performance than when it was preceded by poor per­ formance (Murphy et al., 1985), except where a delay occurred between view­ ing and evaluating the final performance. Both evaluations of performance and estimates of the frequency of good and poor behaviors were susceptible to these contrast effects, which Murphy et al. (1985) attributed to greater attention to be­ havior inconsistent with the initial impression. Recently, Maurer and Alexander (1991) established differential attention to context-discrepant stimuli as the causal factor in contrast effects when different ratees provided the contrast and target stimuli. Using a variant of Newtson, Engquist, and Bois’ (1976) unitizing proce­ dure, they also showed that rating standards did not change in their situation, in contrast to prior findings (Buckley & Armenakis, 1987), and contradicting earli­ er conjectures (Latham, Wexley, & Pursell, 1975). Balzer (1986) showed simi­ lar attentional effects in the recording of information about a single ratee. In particular, incidents contrasting with an initial impression were recorded more frequently. In a different study, Murphy et al. (1986) found that assimilation effects can occur when subjects rate an average performance after viewing either good or poor performance. However, in this study, effects were limited to behavioral fre­ quency estimates, and occurred when rating of performance episodes occurred after a delay. The authors attributed this assimilation effect to the use of an over­ all person impression constructed online. Such an impression may be moderately diagnostic and more accessible than episodic memories after a delay (and clear­ ing of short-term memory). However, Kravitz and Balzer (1992) cast some doubt on the validity of these findings. They pointed out that without knowing how the target stimulus would be rated in the absence of a context manipulation, differ­ ences between good and poor context conditions are hard to interpret. They also questioned the success in establishing an evaluative context itself in the earlier Murphy et al. (1985, 1986) studies. However, Kravitz and Balzer reported a reli­ able contrast effect in their negative context conditions and an assimilation effect

7.

PERFORMANCE APPRAISAL

371

in the positive condition when the comparison was to no-context (or “randomcontext”) control ratings. A connection between evaluative context studies and more fundamental process­ es may be made via Schwarz and Bless’ (1991) “inclusion-exclusion” model. As in Barsalou’s (1987) discussion of category construction and context-dependence, Schwarz and Bless assumed that the representation of an object of judgment may be flexible, and information may be included or excluded depending on context (e.g., the framing or order of questions). The construction of a standard of com­ parison is likewise flexible under some circumstances (Feldman & Lynch, 1988). Thus, Schwarz and Bless (1991) showed that the evaluation of a political party depends on whether a particular well-respected politician is included or exclud­ ed, which depends, in turn, on the format of prior questions. Likewise, the oc­ currence of assimilation or contrast in the appraisal setting seems to depend on whether given performance episodes are included or excluded from the subject’s representation of the target person’s performance or the subject’s evaluative stand­ ard. In the case of contrast effects, the additional processing given to contradic­ tory information once an impression is formed would be expected to influence its inclusion in the representation of a given performance episode. However, to discover a contrast effect in the evaluation of the first performance witnessed (as in Murphy et al., 1986), the rater would have to retrieve and review the ratee’s prior performance in some detail, subsequent to viewing later perform­ ances. Otherwise, no motive could exist for greater attention to behavior dis­ crepant from performances not yet seen. When this is unlikely to occur, neither contrast nor assimilation is found under immediate rating conditions, whereas delay conditions yield assimilation. This is to be expected given the greater volume of information in subsequent episodes and the likelihood of an overall person im­ pression being formed. Purpose of Observation and Appraisal

There are two broad classes of appraisal “purpose” represented in performance appraisal research. The first class is the use to which the appraisal will be put, characterized as “research versus administrative.” The second is more similar to “processing objective” research in social cognition (i.e., the effect of the nature of the decision to be made on the basis of the appraisal). It should be noted here that both topics assume that the objective of any behavioral observation or other information processing that takes place is to perform an appraisal. In contrast, the “observational purpose” label refers to research in which the objective of form­ ing an evaluation of performance is contrasted with some other objectives. Research Versus Administration. This general purpose manipulation, perhaps most important from an applied perspective, is also the simplest in terms of theo­ retical explanations for its effect. Telling raters that their responses are to be used

372

FELDMAN

for “research purposes only” typically produces lower overall evaluations than telling raters their responses will be used to make administrative decisions about ratees (e.g., raise, promotion, etc.). This seems to be a straightforward influence on the rating response—a function of people’s unwillingness to cause negative consequences to others or to experience interpersonal difficulty in the feedback process (Longnecker et al., 1987). An important question for both theoretical and practical purposes has not been addressed, however. What is the effect of expressing this relatively lenient rating on later judgments? One might imagine that the initial judgment would be stored and accessed to form later judgments, which, if made under different circumstances, would not reflect leniency. However, it also might be that memory of the rating response might bias subse­ quent judgments, via “motivated reasoning” processes (Kunda, 1990), integra­ tion of the recalled rating with other stored information, or the relative accessibility of both the original judgment and the rating response. In the latter case, the process would be similar to the later use of impressions based on “to be disregarded” in­ formation (Wyer & Budesheim, 1987; Wyer & Unverzagt, 1985). However, care must be taken to distinguish effects on judgment from subjects’ attempts to main­ tain a socially desirable response consistency. The Use of Appraisal Information. Williams, DeNisi, Blencoe, and Cafferty (1985) divided the use of appraisals into two decision types: “deservingness,” cor­ responding to a rating of the degree to which each ratee is entitled to or in need of some treatment (e.g., raise, training, etc.); and “designation,” choosing one ratee from a group to receive some treatment. Williams et al. (1985) showed that consensus information (comparing different worker’s performances on the same task), distinctiveness information (comparing a single worker’s performance across multiple tasks), and consistency information (comparing a single worker’s per­ formance on a single task across occasions) are used equivalently for the two types of decisions if these criteria are made equally salient to the raters. However, raters seek more distinctiveness information for deservedness ratings than desig­ nation, whereas designation leads more consensus information to be sought. This is understandable in terms of the requirements for each decision; for designa­ tion, consensus is most diagnostic; whereas for deservingness, distinctiveness mat­ ters more. Williams, DeNisi, Meglino, and Cafferty (1986) found that the purpose of appraisal also influences the nature of performance impressions formed when mul­ tiple ratees are evaluated on multiple tasks. They also varied the organization of information presented to raters, blocking by ratees (corresponding to consen­ sus) or tasks (corresponding to distinctiveness). Two days after making initial designation or deservedness decisions, subjects provided overall performance rat­ ings and ratings of four separate task performances for each ratee. The overall performance ratings made by subjects who had initially made deservedness judg­ ments distinguished between three manipulated levels of ratee performance, but

7.

PERFORMANCE APPRAISAL

373

only when information was blocked by persons. In contrast, subjects who initial­ ly had made designation judgments only could differentiate the best overall work­ er from all others, regardless of blocking pattern. No comparable effects on task-specific performance ratings were found. The pattern of results indicated that the nature of the evaluative impression formed by raters depends on their initial processing objectives and the formation of associations between ratees and specific behaviors. That is, the harder it is to construct individuated impressions, the less likely it is that such detail will be available later regardless of its diag­ nosticity. Thus, even though schematic representations of ratees may permit higher accuracy after a delay (presumably by avoiding the dilution effect of irrelevant detail; Tetlock & Boettger, 1989), the nature of the representation constructed seems to be crucial (Murphy & Balzer, 1986). Observational Purpose. The topic of observational purpose concerns the ob­ jective of performance appraisal as opposed to other processing objectives. In everyday encounters, people are unlikely to have appraisal as the sole or dominant focus of interaction. Rather, their objective is to accomplish some work-related or social goal, and if appraisal motives enter at all it is only in the context of attaining or failing to attain the goal. Either the form or the content of informa­ tion processing and the impression formed may differ depending on the orienta­ tion of the observer. Foti and Lord (1987) manipulated subjects’ knowledge of group goals and the purpose of observation (memory vs. impression formation) in subjects who watched a videotaped group interaction. They found that either goal knowledge or memory purpose elicited a “script” organization of memory, whereas impres­ sion formation or no goal knowledge conditions led to the use of a leadership prototype schema, as indicated by memory clustering measures and the pattern of recognition errors. Apparently, impression-formation objectives are default operations in observation, and other organizing schemata are used when they are called for by context (Bargh, 1989). This suggests that observational purpose should have effects on memoryjudgment relationships in performance appraisal similar to those observed in im­ pression formation studies. That is, when the purpose of observation is either to prevent the formation of a spontaneous impression (via the use of the subject’s processing capacity) or to create an impression that is not diagnostic for a later judgment, episodic memory will be searched and relevant incidents retrieved (Hastie & Park, 1986; Lichtenstein & Srull, 1987; Wyer & Srull, 1986). However, when memories are made accessible in other ways, they should be incorporated into the judgment. There are two arguments supporting this prediction. First, specific, relevant behavioral memories are likely to be perceived as generally more diagnostic than summary judgments, and so will be used if they are as ac­ cessible. Second, the norm of conversational relevance, which also applies in experimental settings (Grice, 1975; Schwarz & Bless, 1991), implies that the

374

FELDMAN

subject is expected to use information that the researcher has taken the trouble to request. Woehr and Feldman (1993) found support for this prediction. Subjects given either a learning set or an alternative (“tape quality”) processing set for viewing a videotape of an economics lecture showed substantial memory-judgment corre­ lations, regardless of whether recall was required before or after ratings of the lecturer’s performance. However, when the subjects’ purpose was to evaluate per­ formance, recall-judgment correlations were significant only when recall preceded ratings. Structural equation analyses supported a “biased-memory” model in the evaluation set-rating first condition, and an availability model in the evaluation set-rating last condition. Williams, Cafferty, and DeNisi (1990) reported similar results from a study dealing with the evaluation of multiple task performances by multiple ratees. Subjects either were asked to evaluate performance or determine task difficulty. Order of rating and recall was counterbalanced (in Study 1), as in Woehr and Feldman’s (1993) study. Evaluation objectives led to person-based memory or­ ganization and higher recall. In addition, the recall/rate order increased recall under task difficulty instructions, whereas rate/recall order improved recall un­ der evaluation objective conditions. Furthermore, rating first led to better dis­ crimination between the overall evaluation of the best worker and all others under evaluation objectives, but recall first led to better discrimination when subjects were trying to determine task difficulty. Memory-judgment correlations were significant in task difficulty, but not in evaluation-objective conditions, regard­ less of recall order. A second study showed that, relative to task blocking, person blocking or no organization instructions led to assimilation effects for separate task ratings in the direction of overall performance, but only when an evaluation goal was not salient at the time the information was received. Because all subjects in Study 2 completed recall measures prior to ratings, the results of the two studies in combination suggest that; (a) evaluative person impressions were the source of both overall and task ratings; (b) these impressions may have been weaker and less detailed under instructions to assess task difficulty; (c) forced recall makes more diagnostic information accessible in case detailed impressions are absent, leading to more differential and stereotype accuracy, though not necessarily to improved differential elevation; and (d) the diagnosticity of episodic memory ver­ sus initial impressions for performance and other person impressions deserves further investigation. DeNisi and Peters (1991, 1992) recently extended the study of recall and its organization to field settings with complementary results. Either on-line or memory-based processing can produce equivalent accuracy. Murphy, Philbin, and Adams (1989) varied observational purpose (evaluation vs. learning) and delay (immediate or 1, 3, or 7 days’ delay in rating) for subjects observing a lecture. Evaluative ratings showed no effect of purpose or delay on three accuracy measures (stereotype, differential, or differential elevation).

7.

PERFORMANCE APPRAISAL

375

However, recognition memory depended on both purpose and delay. Stereotype accuracy was higher in the evaluative condition with delays of 3 days or less, but was higher in the learning condition after 7 days. These results were paralleled by a similar relationship on d' or memory sensitivity. Although it seems like content-focused subjects had an accuracy advantage, the recognition test items largely were concerned with material irrelevant to evaluation, but relevant to lec­ ture content (e.g., topical coverage). It may be that content subjects encoded this information in the course of meeting their processing objectives, perhaps in a script schema (Foti & Lord, 1987). It also seems that evaluation-set subjects be­ come increasingly schematic in their responding over time, because episodic memories diagnostic for the recognition task become relatively less accessible. Small variations in observational purpose may influence memory organiza­ tion even under a general evaluative set. Following Cafferty, DeNisi, and Wil­ liams (1986), DeNisi, Robbins, and Cafferty (1989) found that keeping a diary of observed incidents of (simulated) job performance led subjects to organize their behavior memories in ways consistent with the manner in which they were in­ structed to organize their diaries. Furthermore, people preferred to organize di­ aries by ratee, rather than by task, when diary organization was unconstrained. Person organization led to better overall accuracy and task-by-task differentia­ tion (comparable to differential elevation, stereotype accuracy, and differential accuracy). The diary manipulation may be compared to cognitive tuning (Zajonc, 1960), in that the demand to record information in a particular form logically should affect the nature of its elaboration and organization. Without such a de­ mand, less organization (and poorer recall) is to be expected. Diary keeping may be influenced by context, however. Kozlowski and Ford (1991) found that raters’ search for performance information was moderated by their prior information, both amount and direction. More information led to less search unless the information was negative. In that case, more information was sought. As discussed in DeNisi et al. (1989), and as implied by studies cited previ­ ously, information acquisition and diary entry may be influenced by existing im­ pressions and search heuristics. Specific training in observation may be necessary to maximize accuracy. This point is discussed later. Organizational and Interpersonal Context

A rater’s relationships with others and with the organization can have a profound influence on both rating responses and judgment. Features of the organizational context that may drive the rating response typically are discussed under the heading of political influences, as noted earlier. These are important, but are considered more properly under the heading of motivation and attitude-behavior relation­ ships, because they are deliberate behaviors in the service of some consciously held goal. However, one feature of the organizational context may operate on judgment

376

FELDMAN

processes, as well as on the choice of a rating response. Tetlock (1985) called this accountability. It is essentially the operation of contingencies on the judg­ ment process. An accountable rater expects to justify his or her ratings to one or more sources, and to experience consequences of the process. Tetlock made the case that, in the natural environment, decision makers are typically account­ able, and thereby questioned the external validity of laboratory studies of judg­ ment and decision processes. Tetlock attempted to outline the strategies used to satisfy the motives operating in organizational and interpersonal settings. Basically, Tetlock’s (1985) position reflected those of Devine (1989), Gilbert (1989), Johnson and Eagly (1989), Petty and Cacioppo (1986), and Eagly and Chaiken (1993), as well as those of Kunda (1990) and Kruglanski (1989). When accountability exists, the cognitive miser is motivated to engage in controlled, central processing. However, the direction such processing takes depends on the nature of the motive to be satisfied. This concern is similar to the point made by Johnson and Eagly (1989), Eagly and Chaiken (1993), and Kunda (1990); the content of processing depends on the type of involvement and the nature of the issue. Tetlock (1985) differentiated three motives: 1. To enhance one’s social image or identity, similar to value-based involve­ ment. 2. To protect and enhance one’s self-image, as in Eagly and Chaiken’s (1993) self-presentation motive. 3. To respond to externally imposed reinforcement or punishment contingen­ cies, which was the nature of Petty and Cacioppo’s (1986) typical involve­ ment manipulation. As might be expected, any kind of accountability/involvement seems to produce more central processing, with concomitant increases in awareness of one’s reason­ ing. In the system of Feldman and Lynch (1988), accountability/involvement moti­ vates a search for more diagnostic information, rather than reliance on one’s stored impression. Kunda’s (1990) concept of accuracy motivation, Kruglanski’s (1989) “fear of invalidity,” “involvement” in the ELM sense (Petty & Cacioppo, 1986), and the presumed accuracy motives of Devine’s (1989) and Gilbert’s (1989) sub­ jects seem to apply. Accountability need not always increase accuracy. As shown by Tetlock and Boettger (1989), accountable people are motivated to use all avail­ able information, relevant or not; thus a “dilution” effect sometimes may occur. This can be traced to the operation of self-presentation motives or conversational norms (as noted previously; Grice, 1975; Schwarz & Bless, 1991), making all information diagnostic. However, subject-matter expertise would be likely to mitigate the effect due to the (presumed) greater diagnosticity of relevant infor­ mation for the expert (Alba & Hutchinson, 1987; Feldman & Lynch, 1988). However, accountability motivated by other types of involvement may pro­ mote distortion more directly. What Tetlock (1985) called the “rationalization

7.

PERFORMANCE APPRAISAL

3 77

heuristic” also was reflected in Kunda’s (1990) discussion of one’s need to arrive at and justify a desired conclusion. This desire to be seen as a reasonable person may be similar to the motive not to appear easily persuaded, as discussed by Eagly and Chaiken (1993); in the judgment case, recruitment of supportive arguments is the result. In persuasion, increased skepticism occurs. Because, theoretically, the increased accessibility of position-diagnostic arguments or memories is in­ volved in the judgment formation process, one might expect higher memoryjudgment correlations (as well as under accuracy motives), increased polariza­ tion of self-generated judgments over time (Tesser, 1978), and possibly reliance on self-generated memories or arguments (Slusher & Anderson, 1987). Biased attention to confirming evidence also might be expected (Darley & Gross, 1983; Einhorn & Hogarth, 1978). In short, it is entirely possible for people to convince themselves that what they want to believe, or what is politically or interpersonally expedient to believe, is true. Klimoski and Inks (1990) manipulated a form of interpersonal accountability (anticipated feedback to subordinates) and exposure to subordinate’s positive or negative self-assessments. Anticipated feedback led to more extreme ratings in the direction of self-assessments, compared with no-feedback conditions. Although no relevant data were provided, this seems clearly a case of motivated changes in ratings, as discussed by Longnecker et al. (1987), in the face of interpersonal contingencies. Rosen and Fiske (1990) showed that men who anticipated a social or romantic interaction with a woman evaluated her (incompetent) performance higher than men who did not have such expectations. They concluded that an­ ticipating a date makes objective elevation difficult. However, this conclusion was not unequivocal. Their data showed that ratings are influenced by the antici­ pation of interaction. This effect may be due either to an influence on judgment or to an interpersonal tactic such as ingratiation. Although simple liking (as well as stronger affections) may bias judgment in an automatic and unintended fashion (Cardy & Dobbins, 1986), it is also true that such biases may be overcome sim­ ply by asking appropriate questions (Lance, Woehr, & Fisicaro, 1991). Both ac­ countability for accuracy and job knowledge, as well as individual value systems (Devine, 1989), should play a role in the judgment and rating processes. Commitment to a ratee is an important fact of organizational life. “Mentor­ ing” and leader-member exchange relationships (Graen, 1976) might produce an escalation bias in judgment similar to that found in decisionmaking, where additional resources sometimes are allocated to a failing enterprise when the decisionmaker has responsibility for the initial allocation. Bazerman, Beekum, and Schoorman (1982) observed a similar phenomenon in a role-playing performance appraisal study. Subjects presented with data suggesting poor performance by a person for whose hiring they had been “responsible” rated him or her higher than those who were not responsible for the original decision. Schoorman (1988) report­ ed the same effect in real-life appraisals. Managers who were opposed to initial hiring decisions became more negative over time, whereas those who supported

37 8

FELDMAN

the hiring became more positive. However, it is not clear whether self-presentation bias influenced ratings in one or both cases; whether differential attention, en­ coding, or retrieval occurred; or whether affect was integrated with other infor­ mation in the process of judgment construction in one or both studies. Summary

It seems clear, then, that motives affect the amount, kind, and direction of process­ ing, and that more extensive processing does not necessarily lead to more ac­ curate judgments, let alone their expression in ratings. Research is needed on the question of how accountability, in the form of various motives for process­ ing, influences the effort put into judgment tasks, the accessibility and diagnosticity of information encountered or retrieved in the process, and the choice of the rat­ ing response. Memory measures (recall and recognition), measures of self­ generated elaborations, and response time should prove helpful in these studies.

INDIVIDUAL DIFFERENCES

Industrial/organizational psychology traditionally has been concerned as much with individual differences as with situational influences on behavior. The study of performance appraisal is no exception. Although the individual differences ef­ fects studied have not been explored in detailed process terms, the results gener­ ally have been compatible with, and complementary to, constructionist processing models. In some cases, they point to issues of potential relevance for social cog­ nition research that have been overlooked. Cognitive Complexity

Research in cognitive complexity, defined in terms of the number of categoriza­ tions one might use in conceptualizing others (and their relationship to one another), was one of the first theory-relevant individual difference variables to be studied in the domain of performance appraisal. As used in person percep­ tion, cognitive complexity is measured in terms of the categories that are used to encode others in everyday encounters (Feldman & Hilterman, 1975). Schneier (1977) reported that more cognitively complex individuals were more accurate when rating others on appraisal forms that demanded more differentiated judg­ ments. However, this “cognitive compatibility” theory proved difficult to sup­ port. Lahey and Saal (1981) found no effects of complexity on measures of halo, leniency, or range restriction in students’ rating of instructors. Likewise, Bernardin, Cardy, and Carlyle (1982) failed to find any effect of complexity on ac­ curacy, halo, confidence in ratings, or acceptability of rating formats in four separate studies. The problem seems to stem from the assumption that complexity

7.

PERFORMANCE APPRAISAL

379

in the interpersonal domain implies complexity in other domains; as studies of expertise have shown, complexity of categorization and/or reasoning does not necessarily transfer across domains (Alba & Hutchinson, 1987). Also, testing for simple format x complexity interactions has not taken into account the process­ ing effects associated with complexity (more specifically, differentiation and in­ tegration of conceptual systems; Scott, 1969). Research in other domains has supported the idea that relatively permanent knowledge structures can mediate performance judgments in predictable ways (Feldman & Lindell, 1990; Feldman & Lynch, 1988; Jussim, 1991). Observational Skills

Before performance may be appraised accurately, it must be observed accurate­ ly. Both theory- and data-driven processing require input representative of the target person’s behavior. Murphy, Garcia, Kerkar, Martin, and Balzer (1982) reported that accuracy in observation, as reflected in frequency ratings of be­ havior, is related to corresponding accuracy in evaluation. That is, scores on Cronbach’s (1955) accuracy components computed on frequency ratings are correlated positively with the same components computed on evaluative ratings (stereotype accuracy being an exception). Of course, it is possible that both ratings are driven by overall impressions that are formed online (Srull & Wyer, 1989). However, that observational skill does play a role in accuracy was shown by Cardy and Kehoe (1984). Field-independent raters were found to be more accurate than fielddependent raters in both high- and low-memory demand conditions, presumably because the field independents were better able to select the information needed from the surrounding context. Generalized attentional skills are only a part of the story, however. Accessible knowledge structures have automatic attentiondirecting properties; field-independent raters may be better at picking relevant information from context, but both dependents and independents need to know what to look for. In addition, raw observation or recall is not evaluation; any type of appraisal scale requires inference as well as observation, and inferential accuracy is guided by existing cognitive structures or implicit theories (Jussim, 1991; Nathan & Alexander, 1985). Implicit Theories

It has been known for some time that people differ in their implicit theories of the world, including those about other people (Kelly, 1955). The role that chron­ ically accessible or schematic constructs play in perceptions of and affective reac­ tions to self and others recently has been the focus of more intensive study (Bargh, 1989; Borman, 1983; Fong & Markus, 1982; Markus, 1977). Implicit theories, in the form of narrative models, have been found to guide attribution and deci­ sionmaking (Pennington & Hastie, 1992; Read, 1987), and it is not surprising

38 0

FELDMAN

to find evidence of their operation in performance appraisal as well. Using a var­ iation of Kelly’s (1955) methods, Borman (1987) found that, although Army officers largely shared a core set of constructs in their theories of NCO effective­ ness, substantial variation existed in how individuals combined concepts in ap­ praisal judgments. Using similar techniques, Jolly, Reynolds, and Slocum (1988) found that performance constructs were organized hierarchically, with values as the highest level organizing principle. They also discovered substantial individu­ al differences in means-ends orientation. Other methods, such as policy capturing, have been used to uncover individu­ al differences in the weights given different types of appraisal information (Hob­ son, Mendel, & Gibson, 1981; Zedeck & Cafry, 1977), suggesting differences in implicit theories. However, policy capturing is not adequate to describe either the nature of implicit theories or the processing steps used in judgment. Brannick and Brannick (1989) found, as would be expected from the nature of categorybased processing, that nonlinear/noncompensatory strategies are used to make performance judgments. Further, McIntyre (1992) found that different ratees elicit different policies from the same judge, as might be expected if relatively vivid descriptions (which he used) resulted in individuated impression-formation processes. Both of these findings call into serious question the use of linear regression-based models to describe individuals’ judgment policies. However, implicit theories are very plausibly related to accuracy, regardless of whether accuracy is described as agreement with experts, retrieval of objec­ tive performance data, or prediction of behavior. In the first case, shared value systems and the categorical components of implicit theories, as well as judgment standards and inferential heuristics, would lead to similar judgments about in­ herently subjective performances (e.g., art, music, etc., as well as such concepts as “managerial style”; Feldman, 1986b, 1992a). In all three cases, the accessibil­ ity of organizing principles and concepts would promote increased recall, memory organization, and inferential accuracy (Nathan & Alexander, 1985), provided that the rater’s model is a valid one. This is essentially an extension of the findings on the role of expertise in problem-solving and judgment. Expertise is assumed to operate at a process level exactly as do implicit theories. This includes both chronic category accessibility allowing spontaneous online judgments and the abil­ ity to reason in terms of principles, rather than on the basis of the surface fea­ tures of stimuli. In keeping with the earlier arguments, Hauenstein and Alexander (1991) found that normative raters (whose implicit theories were similar to most other subjects’) were more accurate in their rating of a videotaped lecturer’s per­ formance than those with idiosyncratic implicit theories when both were observ­ ing under content-learning instructions. They also found that normative raters exhibited more halo in ratings, presumably due to spontaneous impression for­ mation. Rater intelligence also was found to relate to accuracy, but differently for

7.

PERFORMANCE APPRAISAL

381

different components of accuracy and for normative versus idiosyncratic raters. Elevation scores tended to be most accurate at moderate rater ability levels, es­ pecially within idiosyncratic rater groups, due to the tendency of the most intelli­ gent raters to be too strict and the less intelligent to be too lenient. This appears to be a judgment anchoring effect. For “dimensional accuracy,” comparable to differential accuracy, the more intelligent idiosyncratic and less intelligent norm­ ative raters were most, and equally, accurate. Least accurate were the less intel­ ligent idiosyncratic raters. This is attributed to the previously observed tendency for the lower ability normative raters and higher ability idiosyncratic raters to elaborate more during encoding than other groups (Hauenstein & Lord, 1989). Thus, it appears that a normative (and presumably correct) implicit theory al­ lows one to process information quickly and to achieve at least moderate accura­ cy through inferential reasoning (Funder, 1987; Jussim, 1991). Greater accuracy is possible, but only at the cost of the additional effort required for intentional processing and elaboration. The inferential benefits of shared conceptual systems also were shown by Zalesny and Highhouse (1992). Accuracy in predicting the behavior of a ratee (a stu­ dent teacher) was related to agreement with expert ratings of the target person’s performance, but only on the performance dimension related to the behavior in question. Both types of accuracy were related to the rater’s perceived similarity to the ratee in attitudes toward teaching, suggesting that a common implicit job theory drove both evaluations and the (unscripted) behavior of the ratee. This should not be taken to mean that the benefits of expertise are limited to top-down or theory-driven processing. Kozlowski et al. (1986) and Kozlowski and Kirsch (1987) reported evidence suggesting that more knowledgeable raters formed more accurate individuated impressions of familiar ratees, relying more on halo-generating implicit theories for only the relatively less familiar. Less knowledgeable raters relied on implicit theories for the more familiar ratees, but their ratings of the less familiar were almost random. Thus, the expert was better able to differentiate individuals and used an implicit theory when more diagnos­ tic information was lacking. The less expert rater also relied on theory when specif­ ic knowledge was lacking; but (as mentioned earlier) any schema-based processing requires some information on which to operate. Implicit theories’ inferential functions were illustrated by Martell and Guzzo (1991), who had subjects observe a group before being informed that the group performed either well or poorly. Subjects then rated group performance and com­ pleted a rating-scale recognition measure of effective and ineffective behaviors. Under these conditions (in which feedback-mediated categorization could not in­ fluence attention or encoding), performance cues influenced response bias. That is, subjects were more likely to endorse expectation-consistent behavior under both immediate and delayed rating conditions. Thus, where explicit information was lacking, inference was driven by implicit theory.

382

FELDMAN

Summary

Individual differences have been found to influence performance appraisal via what appears to be two distinct mechanisms. The first is the operation of existing knowledge structures, which serve to direct attention, organize memories, and guide inference in a relatively efficient manner within the domain of their (per­ ceived) relevance. The second is the influence of skills and abilities, enabling the person to attend to relevant information and process it more elaborately. Both presumably lead to greater accuracy, although this conclusion depends in part on the accuracy measure chosen. The individual differences discussed earlier referred to the person’s process­ ing capacity: either the extent of processing resources available (as indexed by skill and ability) or the efficiency with which they are used in a given domain (as indexed by knowledge or value structure elaboration). Individual differences relevant to motivation to process have been neglected. For instance, equalitarian values should promote more extensive processing where stereotyping is possible (but not otherwise), as in Devine’s (1989) research. Need for cognition (Eagly & Chaiken, 1993; Petty & Cacioppo, 1986) should promote more elaborate processing in the same manner as accountability (Tetlock, 1985) or accuracy moti­ vation (Kruglanski, 1989). In short, any chronic motive of the individual should influence both the amount and direction of processing observed (Kunda, 1990). This is an area worthy of future research.

TRAINING FOR PERFORMANCE APPRAISAL

A historical concern of individual/organizational (I/O) psychology, and one that diverges sharply from topics typically of interest to social cognition researchers, is training. Because training is one of the most effective tools in the repertoire of I/O psychology, it is not surprising to find it applied to performance appraisal. However, the issues raised by the study of appraisal training are of more than purely applied interest. Considered as an experimental intervention, training offers an opportunity to develop and test theoretical hypotheses as well. Early research on training was informed by an analytic model of the appraisal process, clearly stated by Borman (1978). In Borman’s view (and many others’; see Feldman, 1981), performance appraisal is, and should be, carried out in three stages: 1. Observation of relevant behavior. 2. Evaluation of each behavior. 3. Weighting and integration of each behavior in terms of each dimension of judgment.

7.

PERFORMANCE APPRAISAL

383

As the earlier data indicated, the appraisal process rarely, if ever, takes this form. Spontaneous judgment, inferences from global categorizations, and the like are more common and may in fact be both accurate and efficient. However, with some exceptions, training research continues to follow this analytic model. As might be expected from researchers’ preoccupation with rating errors such as leniency, halo, and range restriction, early training programs tended to focus on the elimination of these errors via lecture and workshop formats. Such “Rater Error Training” (see D. E. Smith, 1986; Woehr & Huffcut, 1992) was in fact effective in reducing these rating tendencies. Unfortunately, the training was at best ineffective in increasing accuracy, and often reduced it (Bernardin & Buck­ ley, 1981; Bernardin & Pence, 1980; Hedge & Kavanagh, 1988; Woehr & Huffcut, 1992). Basically, some rater error training programs instructed recipients in a new response set, creating ratings with lower intercorrelations and mean values regardless of whether these reflected actual performance. However, some pro­ grams instructed trainees in the avoidance of rating errors without specifying target distributions (and thus, new response biases). These programs produced increased accuracy (Latham, 1986; Woehr & Huffcut, 1992). A second form of training, known variously as “Rater Accuracy Training” or “Frame of Reference” training (Bernardin & Beatty, 1984; Pulakos, 1984, 1986), has a focus more in keeping with social cognitive views of the appraisal process. Basically, this training is aimed at transmitting knowledge of performance dimen­ sions and standards of judgment using a variety of techniques (including lecture, dis­ cussion of behavioral examples, and feedback). Woehr and Huffcut (1992) further subdivided this form of training into two categories. The first is “Performance Dimension Training,” intended to familiarize raters with the rating dimensions prior to observation of behavior. Such training presumably increases the accessi­ bility of the relevant rating concepts, which aids in memory organization and the formation of on-line judgments (Woehr & Feldman, 1993). The second is “Frame-of-Reference” training proper, which involves knowledge of evaluative standards and a more sophisticated theory of performance, as discussed previously. Rater accuracy and frame of reference methods teach categories of behavior and provide representations of ideal, average, and unacceptable exemplars/ideals/ prototypes against which to calibrate judgments. An important feature of such training is the identification of individuals with idiosyncratic performance standards or theories, in order to bring their judgments in line with organizationally norm­ ative standards and values (Bernardin & Beatty, 1984; Hauenstein & Foti, 1989). Concept and standard-focused training increases accuracy of rating, and frame of reference training additionally improves observational accuracy (Woehr & Huff­ cut, 1992). Furthermore, training specific to the format of appraisal (e.g., evalu­ ative ratings or recognition-based responding) adds to accuracy (Pulakos, 1986). Woehr and Huffcut (1992) also differentiated “Behavioral Observation Train­ ing” from rater error and rater accuracy training. The goal of observational training

384

FELDMAN

is to improve the detection of relevant behavior. As discussed earlier, those more able to detect relevant behavior in informationally noisy environments logically should be better equipped to make appraisal judgments (Hedge & Kavanagh, 1988; Thornton & Zorich, 1980). This is in fact what was found; observational train­ ing led to increased rating and recognition accuracy. Furthermore, combinations of training types seemed to be effective in increasing accuracy and reducing rater errors (Hedge & Kavanagh, 1988; Woehr & Huffcut, 1992). Summary

Although the training studies conducted to date generally were conceptualized in cognitive terms, they had an almost exclusive focus on “effectiveness,” neglect­ ing tests of process hypotheses. An exception is the work of Hauenstein and Foti (1989), which examined the content of normative and idiosyncratic rater perform­ ance schemata and showed that these could be identified via sensitivity and response-threshold analyses. However, tests of the process by which training is presumed to work are still needed. Raters who are trained in normative perform­ ance theories should show changes in response time, memory organization, memory-judgment correlations, context independence of categories, sensitivity in recognition memory, and freedom from affective/stereotypic biases consistent with increases in the chronic accessibility of normative categories. Furthermore, chronic accessibility and context independence should mediate increases in ac­ curacy. Another profitable line of investigation would be the examination of train­ ing techniques in terms of both mediating cognitive variables and accuracy. Cognitive feedback (Balzer, Doherty, & O’Connor, 1989) and the elaboration of examples (characteristic of learning that produces expertise; Chi, Lewis, Reimann, & Glaser, 1989) are two training factors that deserve investigation. Field studies of training effects are certainly possible, and would add an important dimen­ sion to both social cognition and organizational research.

ISSUES AND DIRECTIONS FOR FUTURE RESEARCH

Compared with the range of process hypotheses tested in the social cognition liter­ ature, research in performance appraisal has been relatively narrowly focused. Compared with the range of settings and contexts studied in performance appraisal, social cognition research likewise has been narrowly focused. Each may benefit from increased breadth. For example, the study of context effects and motivated processing is an area of interest to both applied and theoretical researchers. Affective context, as one instance, is just beginning to be studied in the appraisal arena. Sinclair (1988) showed that the effects of mood on appraisal were similar to those found in other domains; people in negative moods processed information more analytically and

7.

PERFORMANCE APPRAISAL

385

were more accurate than those people in positive moods. Factors in real-world settings, such as job satisfaction, may influence chronic mood, and thus the use of overall impressions and heuristics (vs. the retrieval of specific episodes) to construct appraisals. The nature of accountability (e.g., punishment vs. reinforce­ ment contingencies) also should influence processing strategies, as would the na­ ture of the rater’s involvement in the process (contingency based, value based, or self-presentation based; Eagly & Chaiken, 1993; Kunda, 1990). Because con­ ditions in the natural environment are much more powerful than can be manipu­ lated, ethically or logistically, in the laboratory, theory-based field research should be informative about the (qualitative and quantitative) nature of judgment processes. Studies of the accuracy and validity of appraisals may both inform and be in­ formed by social cognition. For instance, Zalesny and Highhouse (1992) sug­ gested that accurate behavioral prediction depends on a shared model of the world (or at least a specific part of it) that guides both perceptions and behavior. The organization and operation of these structures is of both practical and theoretical interest (Bargh, 1989; Jussim, 1991; Murphy & Medin, 1985). Research on the validity of appraisal judgments as predictors of future performance and behavior may test cognitive models as well as provide important information for person­ nel decisions. Recent theoretical advances should rekindle interest in appraisal format ef­ fects. Both Landy and Farr (1980) and Feldman (1981) recommended against appraisal format research on the grounds that the purely empirical approach of the 1960s and 1970s had not produced any advances in either practice or theoret­ ical knowledge. Recent research on constructive processes in survey respond­ ing, coupled with theoretical developments (Feldman & Lynch, 1988; Schwarz & Bless, 1991; Tourangeau & Rasinski, 1988), provide a more fertile ground for this topic. For example, questions of framing effects on memory retrieval versus the use of overall evaluations may be addressed in situations that are con­ sequential and involving for participants. Woehr and Feldman (1993) and Wil­ liams et al. (1990) already demonstrated that question format influences the use of stored information in judgment, but many questions (e.g., biased retrieval or weighting) remain. The Schwarz and Bless (1991) inclusion/exclusion model pro­ vides several testable hypotheses. One intriguing question is the degree to which the format of appraisal instruments may induce elaboration, with subsequent ef­ fects on memory content, interpersonal behavior, and, ultimately, the performance of the ratee in question (Feldman, 1986a; Jussim, 1986, 1991; Slusher & Ander­ son, 1987). A final and very broad question concerns the development of entire appraisal systems. At present, two contending viewpoints exist: the analytic (Bernardin, 1992; Kane & Kane, 1992) and the nonanalytic (DeNisi, 1992; Feldman, 1992a). The former attempts to improve reliability, validity, and accuracy by using mechanical, analytic weighting, and integration of behavioral frequency infor­

386

FELDMAN

mation. The latter proposes to reach the same ends by engineering appraisal for­ mat, content, and training to promote accurate, well-calibrated, spontaneous per­ formance judgments (Williams, 1992). Each depends on a number of generali­ zations from the social cognition literature (e.g., the degree to which behavioral frequency information is attended to, encoded, and stored automatically and retrieved in an unbiased fashion; Kane & Kane, 1992). The process of designing and implementing such systems provides an opportunity for testing the circum­ stances under which their guiding assumptions are valid, and for testing the process models underlying each. ACKNOWLEDGMENTS

Many thanks are due to Linda Hoopes for her unfailing encouragement, as well as her valuable suggestions; to Bob Wyer and Thom Srull, for extraordinary editorial assistance, patience, and forbearance; and to colleagues and students too numerous to list (but you know who you are), for stimulating argument and Devil’s Advocacy. REFERENCES Alba, J. W ., & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal o f Consumer Research, 13, 411-454. Arvey, R. D., Bouchard, T. J., Jr., Segal, N. L., & Abraham, L. M. (1989). Job satisfaction: En­ vironmental and genetic components. Journal o f Applied Psychology, 74, 187-192. Atkin, R. S., & Conlon, E. J. (1977). Behaviorally anchored rating scales: Some theoretical issues (Working Paper No. 77-75-76). Pittsburgh, PA: Carnegie-Mellon University, Graduate School of Industrial Administration. Austin, J. T ., & Villanova, P. (1992). The criterion problem: 1917-1992. Journal o f Applied P sy­ chology, 77, 836-874. Balzer, W. K. (1986). Biases in the recording of performance related information: The effects of initial impression and centrality of the appraisal task. Organizational Behavior and Human D eci­ sion Processes, 37, 329-347. Balzer, W. K., Doherty, M. E., & O’Connor, R., Jr. (1989). The effects of cognitive feedback on performance. Psychological Bulletin, 106, 410-433. Balzer, W. K., & Sulsky, L. M. (1992). Halo and performance appraisal research: A critical exami­ nation. Journal o f A pplied Psychology, 77, 975-985. Banks, C. G., & Murphy, K. R. (1985). Toward narrowing the research-practice gap in performance appraisal. Personnel Psychology, 38, 335-345. Bargh, J. A. (1984). Automatic and conscious processing of social information. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 3, pp. 1-44). Hillsdale, NJ: Lawrence Erlbaum Associates. Bargh, J. A. (1989). Conditional automaticity: Varieties of automatic influence in social perception and cognition. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 3-51). New York: Guilford. Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and social perception: The influence of trait information presented outside of conscious awareness on impression forma­ tion. Journal o f Personality and Social Psychology, 43, 437-449.

7.

PERFORMANCE APPRAISAL

387

Barsalou, L. W. (1987). The instability of graded structure: Implications for the nature of concepts. In U. Neisser (Ed.), Concepts reconsidered: The ecological and intellectual bases o f categories (pp. 101-140). Cambridge: Cambridge University Press. Bazerman, M. H., & Atkin, R. S. (1978). Performance appraisal: An information-processing and attributional perspective (Working Paper No. 25-78-79). Pittsburgh, PA: Camegie-Mellon Univer­ sity, Graduate School of Industrial Administration. Bazerman, M. H., Beekum, R., & Schoorman, F. (1982). Performance evaluation in a dynamic context: A laboratory study of the impact of a prior commitment to the ratee. Journal o f A pplied Psycholo­ gy, 67, 873-876. Becker, B., & Cardy, R. (1986). Influence of halo error on appraisal effectiveness: A conceptual and empirical reconsideration. Journal o f Applied Psychology, 71, 662-671. Bern, S. L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354-364. Bemmels, B. (1988). Gender effects in discipline arbitration: Evidence from British Columbia. Academy o f Management Journal, 31, 699-706. Bemardin, H. J. (1992). An “analytic” framework for customer-based performance content develop­ ment. Human Resource Management Review, 2, 81-102. Bemardin, H. J., & Beatty, R. W. (1984). Performance appraisal: Human behavior at work. Boston: Kent. Bernardin, H. J., & Buckley, M. R. (1981). A consideration of strategies in rater training. Academy o f Management Review, 6, 205-212. Bernardin, H. J., Cardy, R. L., & Carlyle, J. J. (1982). Cognitive complexity and appraisal effec­ tiveness: Back to the drawing board? Journal o f A pplied Psychology, 67, 151-160. Bemardin, H. J., & Pence, E. C. (1980). Effects of rater training: Creating new response sets and decreasing accuracy. Journal o f Applied Psychology, 65, 60-66. Billman, D., & Knutson, J. F. (1990, November). Systematic versus orthogonal sets o f correlations benefit unsupervised learning. Paper presented at the meeting of the Psychonomic Society, New Orleans, LA. Borman, W. C. (1978). Exploring the upper limits of reliability and validity in job performance rat­ ings. Journal o f Applied Psychology, 63, 135-144. Borman, W. C. (1983). Implications of personality theory and research for the rating of work perfor­ mance in organizations. In F. Landy, S. Zedeck, & J. Cleveland (Eds.), Performance measure­ ment and theory (pp. 127-165). Hillsdale, NJ: Lawrence Erlbaum Associates. Borman, W. C. (1987). Personal constructs, performance schemata, and “folk theories” of subor­ dinate effectiveness: Explorations in an army officer sample. Organizational Behavior and Hu­ man Decision Processes, 40, 307-322. Borman, W. C., White, L. A., Pulakos, E.D ., & Oppler, S. H. (1991). Models of supervisory job performance ratings. Journal o f Applied Psychology, 76, 863-872. Bottger, P. C., & Yetton, P. W. (1988). An integration of process and decision scheme explanations of group problem solving performance. Organizational Behavior and Human Decision Processes, 43, 234-249. Brannick, M ., & Brannick, J. (1989). Nonlinear and noncompensatory processes in performance evaluation. Organizational Behavior and Human Decision Processes, 44, 97-122. Brewer, M. B. (1988). A dual process model of impression formation. In T. K. Srull & R. S. Wyer, Jr. (Eds.), Advances in social cognition (Vol. 1, pp. 1-36). Hillsdale, NJ: Lawrence Erlbaum As­ sociates. Brewer, W. F., & Nakamara, G. V. (1984). The nature and functions of schemas. In R. S. Wyer, Jr., & T. K. Srull (Eds.), The handbook o f social cognition (Vol. 1, pp. 119-160). Hillsdale, NJ: Lawrence Erlbaum Associates. Buckley, R. M., & Armenakis, A. A. (1987). Detecting scale recalibration in survey research. Group and Organization Studies, 12, 464-481. Cafferty, T. P., DeNisi, A. S., & Williams, K. J. (1986). Search and retrieval patterns for perform­ ance information: Effects on evaluation of multiple targets. Journal o f Personality and Social Psy­ chology, 50, 676-683.

388

FELDMAN

Cardy, R. L., & Dobbins, G. H. (1986). Affect and appraisal accuracy: Liking as an integral dimen­ sion in evaluating performance. Journal o f Applied Psychology, 71, 672-678. Cardy, R. L., & Kehoe, J. F. (1984). Rater selective attention ability and appraisal effectiveness: The effect of cognitive style on the accuracy of differentiation among ratees. Journal o f Applied Psychology, 69, 589-594. Chi, M. T. H., Lewis, M. W ., Reimann, P., & Glaser, R. (1989). Self-explanations: How students study and use examples in learning to solve problems. Cognitive Science, 13, 145-182. Christensen-Szalanski, J. J. J., & Beach, L. R. (1984). The citation bias: Fad and fashion in the judgment and decision literature. American Psychologist, 39, 75-78. Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative conduct: A theoretical refinement and reevaluation of the role of norms in human behavior. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 202-235). New York: Academic Press. Cleveland, J. N., & Landy, F. J. (1983). The effects of person and job stereotypes on two personnel decisions. Journal o f Applied Psychology, 68, 609-619. Cleveland, J. N., Murphy, K. R., & Williams, R. E. (1989). Multiple uses of performance appraisal: Prevalence and correlates. Journal o f Applied Psychology, 74, 130-135. Cooper, W. H. (1981). Ubiquitous halo. Psychological Bulletin, 90, 218-244. Cronbach, L. J. (1955). Processes affecting scores on understanding of others and assumed “similar­ ity.” Psychological Bulletin, 52, 177-193. Darley, J. M., & Gross, P. H. (1983). A hypothesis-confirming bias in labeling effects. Journal o f Personality and Social Psychology, 44, 20-33. Davis-Blake, A., & Pfeffer, J. (1989). Just a mirage: The search for dispositional effects in organiza­ tional research. Academy o f Management Review, 14, 385-400. Deaux, K. (1984). Sex and gender. Annual Review o f Psychology, 36, 49-81. DeNisi, A. S. (1992). Why analytic approaches to performance appraisal cannot work. Human Resource Management Review, 2, 71-80. DeNisi, A. S., Cafferty, T. P., & Meglino, B. M. (1984). A cognitive view of the performance appraisal process: A model and research propositions. Organizational Behavior and Human P er­ form ance, 33, 360-396. DeNisi, A. S., & Peters, L. H. (1991, August). Memory reorganization and perform ance appraisal: A field experiment. Paper presented at the meeting of the Academy of Management, Miami, FL. DeNisi, A. S., & Peters, L. H. (1992, April). D iary keeping and the organization o f information in memory: A field extension. Paper presented at the meeting of the Society for Industrial-Organiza­ tional Psychology, Montreal, Quebec, Canada. DeNisi, A. S., Robbins, T., & Cafferty, T. P. (1989). Organization of information used for perform­ ance appraisals: Role of diary-keeping. Journal o f Applied Psychology, 74, 124-129. DeNisi, A. S., & Williams, K. J. (1988). Cognitive approaches to performance appraisal. In G. R. Ferris & K. M. Rowland (Eds.), Research in personnel and human resources management (Vol. 6, pp. 109-155). Greenwich, CT: JAI. Devine, P. G. (1989). Automatic and controlled processes in prejudice: The role of stereotypes and personal beliefs. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 181-212). Hillsdale, NJ: Lawrence Erlbaum Associates. Dipboye, R. L. (1985). Some neglected variables in research on discrimination in appraisals. Acade­ my o f Management Review, 10, 116-127. Dobbins, G. H ., Cardy, R. L., & Truxillo, D. M. (1988). Effects of purpose of appraisal and in­ dividual differences in stereotypes of women on sex differences in performance ratings: A labora­ tory and field study. Journal o f Applied Psychology, 73, 551-558. Downs, A. C., & Lyons, P. M. (1991). Natural observations of the links between attractiveness and initial legal judgments. Personality and Social Psychology Bulletin, 17, 541-547. Dunnette, M. D ., & Hough, L. (1991). Handbook o f industrial and organizational psychology (2nd ed.) Palo Alto, CA: Consulting Psychologist’s Press.

7.

PERFORMANCE APPRAISAL

3 89

Eagly, A. H., Ashmore, R. D., Makhijani, M. G., & Longo, L. C. (1991). What is beautiful is good, b u t . . . : A meta-analytic review of research on the physical attractiveness stereotype. Psy­ chological Bulletin, 110, 109-121. Eagly, A. H., & Chaiken, S. (1993). The psychology o f attitudes. Fort Worth, TX: Harcourt Brace Jovanovich. Eagly, A. H., Makhijani, M. G., & Klonsky, B. G. (1992). Gender and the evaluation of leaders: A meta-analysis. Psychological Bulletin, 111, 3-27. Eden, D., & Shani, A. B. (1982). Pygmalion goes to boot camp: Expectancy, leadership, and trainee performance. Journal o f Applied Psychology, 67, 194-199. Einhom, H. J., & Hogarth, R. M. (1978). Confidence in judgment: Persistence of the illusion of validity. Psychological Review, 85, 395-416. Fazio, R. H. (1990). Multiple processes by which attitudes guide behavior: The MODE model as an integrative framework. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 75-110). New York: Academic Press. Feldman, J. M. (1972). Stimulus characteristics and subject prejudice as determinants of stereotype attribution. Journal o f Personality and Social Psychology, 21, 333-340. Feldman, J. M. (1981). Beyond attribution theory: Cognitive processes in performance appraisal. Journal o f A pplied Psychology, 66, 127-148. Feldman, J. M. (1986a). A note on the statistical correction of halo error. Journal o f Applied P sy­ chology, 71, 173-176. Feldman, J. M. (1986b). Instrumentation and training for performance appraisal: A perceptual-cognitive viewpoint. In K. Rowland & G. Ferris (Eds.), Research in personnel and human resources manage­ ment (Vol. 4, pp. 49-100). Greenwich, CT: JAI. Feldman, J. M. (1991). Social cognition, reality, and job perception. In L. W. Hong (Ed.), P erspec­ tives on judgm ent and decision making. Metuchen, NJ: Scarecrow Press. Feldman, J. M. (1992a). The case for non-analytic performance appraisal. Human Resource Manage­ ment Review, 3, 9-35. Feldman, J. M. (1992b). Constructive processes as a source of context effects in survey research: Explorations in self-generated validity. In N. Schwarz & S. Sudman (Eds.), Context effects in so­ cial and psychological research (pp. 49-62). New York: Springer-Verlag. Feldman, J. M ., & Hilterman, R. J. (1975). Stereotype attribution revisited: The role of stimulus characteristics, racial attitude, and cognitive differentiation. Journal o f Personality and Social P sy­ chology, 31, 1177-1188. Feldman, J. M ., & Hilterman, R. J. (1977). Sources of bias in performance evaluation: Two experi­ ments. International Journal o f Intercultural Relations, 1, 35-57. Feldman, J. M ., & Lindell, M. K. (1990). On rationality. In I. Horowitz (Ed.), Organization and decision theory (pp. 93-166). Boston, MA: Kluwer Academic. Feldman, J. M., & Lynch, J. G., Jr. (1988). Self-generated validity and other effects of measure­ ment on belief, attitude, intention and behavior. Journal o f A pplied Psychology, 73, 421-435. Fisicaro, S. A. (1988). A reexamination of the relation between halo error and accuracy. Journal o f Applied Psychology, 73, 239-244. Fiske, S. T., Bersoff, D. N., Borgida, E., Deaux, K., & Heilman, M. E. (1991). Use of sex stereotyping research in Price Waterhouse vs. Hopkins. American Psychologist, 46, 1049-1060. Fiske, S. T ., & Neuberg, S. L. (1990). A continuum of impression formation, from category-based to individuating processes: Influence of information and motivation on attention and interpreta­ tion. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 1-74). New York: Academic Press. Fiske, S. T., & Pavelchak, N. A. (1986). Category-based vs. piecemeal affective responses: De­ velopments in schema-triggered affect. In R. M. Sorrentino & E. T. Higgins (Eds.), The hand­ book o f motivation and cognition: Foundations o f social behavior (pp. 167-203). New York: Guilford. Fong, G. T ., & Markus, H. (1982). Self-schemas and judgments about others. Social Cognition, 1, 191-204.

390

FELDMAN

Ford, J. K., Kraiger, K., & Schectman, S. L. (1986). Study of race effects in objective indices and subjective evaluations of performance: A meta-analysis of performance criteria. Psychological Bulle­ tin, 79, 330-337. Foti, R., & Lord, R. G. (1987). Prototypes and scripts: The effects of alternative methods of processing information on rating accuracy. Organizational Behavior and Human Decision Processes, 39, 318-340. Funder, D. C. (1987). Errors and mistakes: Evaluating the accuracy of social judgment. Psychologi­ cal Bulletin, 101, 75-90. Gilbert, D. T. (1989). Thinking lightly about others: Automatic components of the social inference process. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 189-211). New York: Guilford. Glick, P., Zion, C., & Nelson, C. (1988). What mediates sex discrimination in hiring decisions? Journal o f Personality and Social Psychology, 55, 178-186. Graen, G. (1976). Role making processes within complex organizations. In M. D. Dunnette (Ed.), Handbook o f industrial and organizational psych ology, (ls te d ., pp. 1201-1248). Chicago: Rand McNally. Graen, G. B., & Scandura, T. A. (1986). A theory of dyadic career reality. In K. M. Rowland & G. R. Ferris (Eds.), Research in personnel and human resources management (Vol. 4, pp. 147-182). Greenwich, CT: JAI. Grice, H. P. (1975). Logic and conversation. In P. Cole & L. L. Morgan (Eds.), Syntax and seman­ tics, 3: Speech acts (pp. 41-58). New York: Academic Press. Hamilton, D. L. (1979). A cognitive-attributional analysis of stereotyping. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 12, pp. 53-82). New York: Academic Press. Harvey, R. J. (1982). The future of partial correlations as a means to reduce halo in performance ratings. Journal o f A pplied Psychology, 67, 171-176. Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the judgment task is memory-based or on-line. Psychological Review, 93, 258-268. Hauenstein, N. M. A., & Alexander, R. A. (1991). Rating ability in performance judgments: The joint influence of implicit theories and intelligence. Organizational Behavior and Human Decision Processes, 50, 300-323. Hauenstein, N. M. A., & Foti, F. J. (1989). From laboratory to practice: Neglected issues in im­ plementing frame-of-reference rater training. Personnel Psychology, 42, 359-378. Hauenstein, N. M. A., & Lord, R. G. (1989, April). The effects o f individual differences on the encoding o f performance information . Paper presented at the meeting of the Society for IndustrialOrganizational Psychology, Boston, MA. Hedge, J., & Kavanagh, M. (1988). Improving the accuracy of performance evaluations: Com­ parison of three methods of performance appraiser training. Journal o f Applied Psychology, 73, 68-73. Heilman, M. E., & Herlihy, J. M. (1984). Affirmative action, negative reaction? Some moderating conditions. Organizational Behavior and Human Performance, 33, 204-213. Heilman, M. E., Lucas, J., & Kaplow, S. (1990). Self-derogating consequences of sex-based preferential selection: The moderating role of initial self-confidence. Organizational Behavior and Human D e­ cision Processes, 46, 202-216. Heilman, M. E., & Martell, R. F. (1986). Exposure to successful women: Antidote to sex discrimi­ nation in applicant screening decisions? Organizational Behavior and Human Decision Processes, 37, 376-390. Heilman, M. E., Martell, R. F., & Simon, M. (1988). The vagaries of sex bias: Conditions regulat­ ing the underevaluation, equivaluation and overevaluation of female job applicants. Organization­ al Behavior and Human Decision Processes, 41, 98-110. Heilman, M. E., & Stopeck, M. H. (1985a). Attractiveness and corporate success: Different causal attributions for males and females. Journal o f A pplied Psychology, 70, 379-388.

7.

PERFORMANCE APPRAISAL

391

Heilman, M. E., & Stopeck, M. H. (1985b). Being attractive, advantage or disadvantage? Performancebased evaluations and recommended personnel actions as a function of appearance, sex, and job type. Organizational Behavior and Human Decision P rocesses, 35, 202-212. Higgins, E. T. (1989). Self-discrepancy theory: What patterns of self-beliefs cause people to suffer? In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22, pp. 93-136). New York: Academic Press. Hobson, C. J., Mendel, R. M ., & Gibson, F. W. (1981). Classifying performance appraisal criteria. Organizational Behavior and Human Performance, 28, 164-188. House, R. J., & Singh, J. V. (1987). Organizational behavior: Some new directions for I/O Psychol­ ogy. Annual Review o f Psychology, 40, 327-351. Hulin, C. L. (1982). Some reflections on general performance dimensions and halo rating error. Journal o f Applied Psychology, 67, 165-170. Iffaldano, M ., & Muchinsky, P. (1985). Job satisfaction and job performance: A meta-analysis. P sy­ chological Bulletin, 97, 251-273. Ilgen, D. R., & Favero, J. L. (1985). Limits in generalization from psychological research on per­ formance appraisal processes. Academ y o f Management Review, 10, 311-321. Ilgen, D. R., & Feldman, J. M. (1983). Performance appraisal: A process focus. In L. Cummings & B. Staw (Eds.), Research in organizational behavior (Vol. 5, pp. 141-197). Greenwich, CT: JAI. Ilgen, D. R., & Klein, J. J. (1989). Organizational behavior. Annual Review o f Psychology, 40, 327-351. Ilgen, D. R., & Youtz, M. A. (1986). Factors affecting the evaluation and development of minorities in organizations. In K. W. Rowland & E. R. Ferris (Eds.), Research in personnel and human resources management (Vol. 4, pp. 307-338). Greenwich, CT: JAI. Isenberg, D. J. (1986). Group polarization: A critical review and meta-analysis. Journal o f Person­ ality and Social Psychology, 50, 1141-1151. James, L. R., & Tetrick, L. E. (1986). Confirmatory analytic tests of three causal models relating job perceptions to job satisfaction. Journal o f Applied Psychology, 71, 77-82. Johnson, B. T., & Eagly, A. H. (1989). The effects of involvement on persuasion: A meta-analysis. Psychological Bulletin, 106, 290-314. Jolly, J., Reynolds, T ., & Slocum, J. (1988). Application of the means-ends theoretic for under­ standing the cognitive bases of performance appraisal. Organizational Behavior and Human D eci­ sion Processes, 41, 153-179. Jones, E. E., & Davis, K. E. (1965). From acts to dispositions: The attribution process in person perception. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2, pp. 133-176). New York: Academic Press. Jussim, L. (1986). Self-fulfilling prophecies: A theoretical and integrative review. Psychological Review, 93, 429-455. Jussim, L. (1991). Social perception and social reality: A reflection-construction model. Psychologi­ cal Review, 98, 54-73. Kane, J. S., & Kane, K. F. (1992). The analytic framework: The most promising approach for the advancement of performance appraisal. Human Resource Management Review, 2, 37-70. Kanfer, R. (1991). Motivation theory and industrial/organizational psychology. In M. D. Dunnette & L. Hough (Eds.), Handbook o f industrial and organizational psychology (2nd ed.). Vol. 1: The­ ory in industrial and organizational psychology (pp. 75-170). Palo Alto, CA: Consulting Psychol­ ogist’s Press. Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitudetreatment interaction approach to skill acquisition [Monograph]. Journal o f Applied Psychology, 74, 657-669. Kelley, H. H. (1967). Attribution theory in social psychology. In D. Levine (Ed.), Nebraska sympo­ sium on motivation (pp. 192-238). Lincoln, NE: University of Nebraska Press. Kelly, G. A. (1955). The psychology o f personal constructs. New York: Norton.

392

FELDMAN

Kenny, D. A. (1991). A general model of consensus and accuracy in interpersonal perception. Psy­ chological R eview, 98, 155-163. Klayman, J., & Ha, Y. W. (1987). Confirmation, disconfirmation, and information in hypothesis testing. Psychological Review, 94, 211-228. Klimoski, R., & Inks, L. (1990). Accountability forces in performance appraisal. Organizational Behavior and Human Decision Processes, 45, 194-208. Kozlowski, S. W. J., & Ford, J. K. (1991). Rater information acquisition processes: Tracing the effects of prior knowledge, performance level, search constraint and memory demand. Organiza­ tional Behavior and Human Decision Processes, 49, 282-301. Kozlowski, S. W. J., & Kirsch, M. P. (1987). The systematic distortion hypothesis, halo, and ac­ curacy: An individual-level analysis. Journal o f Applied Psychology, 72, 252-261. Kozlowski, S. W. J., Kirsch, M. P., & Chao, G. T. (1986). Job knowledge, ratee familiarity, and halo error: An exploration. Journal o f A pplied Psychology, 71, 45-49. Kozlowski, S. W. J., & Mongillo, M. (1992). The nature of conceptual similarity schemata: Exami­ nation of some basic assumptions. Personality and Social Psychology Bulletin, 18, 88-95. Kraiger, K., & Ford, J. K. (1985). A meta-analysis of ratee race effects in performance ratings. Journal o f A pplied Psychology, 70, 56-65. Kravitz, D. A., & Balzer, W. K. (1992). Context effects in performance appraisal: A methodologi­ cal critique and empirical study. Journal o f A pplied Psychology, 53, 14-29. Kruglanski, A. W. (1989). The psychology of being “right”: The problem of accuracy in social per­ ception and cognition. Psychological Bulletin, 106, 395-409. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480-498. Lahey, M. A., & Saal, F. E. (1981). Evidence incompatible with a cognitive compatibility theory of rating behavior. Journal o f A pplied Psychology, 66, 706-715. Lance, C. E., Woehr, D. J., & Fisicaro, S. A. (1991). Cognitive categorization processes in per­ formance evaluation: Confirmatory tests of two models. Journal o f Organizational Behavior, 12, 1- 20 . Landy, F. J., & Farr, J. L. (1980). Performance rating. Psychological Bulletin, 87, 72-107. Landy, F. J., & Farr, J. L. (1983). The measurement o f work performance. New York: Academic Press. Latham, G. P. (1986). Job performance and appraisal. In C. L. Cooper & I. Robertson (Eds.), Inter­ national review o f industrial and organizational psychology. London: Wiley. Latham, G. P., Wexley, K. N., & Pursell, E. D. (1975). Training managers to minimize rating er­ rors in the observation of behavior. Journal o f A pplied Psychology, 60, 550-555. Laughlin, P. R. (1988). Collective induction: Group performance, social combination processes, and mutual minority and majority influence. Journal o f Personality and Social Psychology, 58, 254-267. Lawrence, B. (1988). New wrinkles in the theory of age: Demography, norms, and performance ratings. Academy o f Management Journal, 37, 309-337. Lee, J. A., & Clemons, T. (1985). Factors affecting employment decisions about older workers. Journal o f Applied Psychology, 70, 785-788. LeVine, R. A., & Campbell, D. T. (1972). Ethnocentrism. New York: Wiley. Lichtenstein, M ., & Srull, T. K. (1987). Processing objectives as a determinant of the relationship between recall and judgment. Journal o f Experimental Social Psychology, 23, 93-118. Locke, E. A. (1976). The nature and causes of job satisfaction. In M. D. Dunnette (Ed.), Hand­ book o f industrial and organizational psychology (1st ed., pp. 1297-1350). Chicago: Rand McNally. Locke, E. A., & Latham, G. P. (1990). A theory o f goal setting and task performance. Englewood Cliffs, NJ: Prentice-Hall. Locksley, A., Hepburn, C., & Ortiz, U. (1982). Social stereotypes and judgments of individuals: An instance of the base-rate fallacy. Journal o f Experimental Social Psychology, 18, 23-42. Longnecker, C. O., Sims, H. P., Jr., & Gioia, D. A. (1987). Behind the mask: The politics of em­ ployee appraisal. Academy o f Management Executive, 1, 183-193.

7.

PERFORMANCE APPRAISAL

3 93

Lopes, L. L. (1987). Between hope and fear: The psychology of risk. In L. Berkowitz (Ed.), Ad­ vances in experimental social psychology (Vol. 20, pp. 255-296). New York: Academic Press. Lord, R. G. (1985a). An information-processing approach to social perceptions, leadership and be­ havioral measurement in organizations. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 7, pp. 87-128). Greenwich, CT: JAI. Lord, R. G. (1985b). Accuracy in behavioral measurement: An alternative definition based on rater’s cognitive schema and signal detection theory. Journal o f A pplied Psychology, 70, 66-71. Lord, R. G., Foti, R. J., & Phillips, J. S. (1982). A theory of leadership categorization. In J. G. Hunt & C. Schriesheim (Eds.), Leadership : Beyond establishment views (pp. 104-121). Carbondale, IL: Southern Illinois University Press. Lord, R. G., & Maher, K. J. (1991). Cognitive theory in industrial/organizational psychology. In M. D. Dunnette & L. Hough (Eds.), Handbook o f industrial and organizational psychology (2nd ed.): Vol. 1. Theory in industrial and organizational psychology. Palo Alto, CA: Consulting Psy­ chologist’s Press. MacKenzie, S. B., Podsakoff, P. M ., & Fetter, R. (1991). Organizational citizenship behavior and objective productivity as determinants of managerial evaluations of salesperson’s performance. Or­ ganizational Behavior and Human Decision Processes, 50, 123-150. Markus, H. (1977). Self-schemata and processing information about the self. Journal o f Personality and Social Psychology, 13, 520-535. Martell, R. F., Lane, D. M ., & Willis, C. E. (1992, August). Demonstrating the danger o f using *variance explained” to assess the practical significance o f research findings: A little sex bias can hurt women a lot. Paper presented at the meeting of the American Psychological Association,

Washington, DC. Martell, R. F., & Guzzo, R. A. (1991). The dynamics of implicit theories of group performance: When and how do they operate? Organizational Behavior and Human Decision Processes, 50, 51-74. Maurer, T. J., & Alexander, R. (1991). Contrast effects in behavioral measurement: An investiga­ tion of alternative process explanations. Journal o f Applied Psychology, 76, 3-10. Maurer, T. J., & Taylor, M. A. (in press). Is sex by itself enough? An exploration of gender bias issues in performance appraisal. Organizational Behavior and Human Decision Processes. McIntyre, M. D. (1992). The impact o f targets on the information processing o f judges. Unpublished masters’ thesis, University of Tennessee, Knoxville. Mitchell, T. R., Green, S. G., & Wood, R. E. (1981). An attributional model of leadership and the poor-performing subordinate: Development and validation. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 3, pp. 197-234). Greenwich, CT: JAI. Mobley, W. H., Griffeth, R. W ., Hand, H. H ., & Meglino, B. M. (1979). Review and conceptual analysis of the employee turnover process. Psychological Bulletin, 86, 493-522. Murphy, G. L., & Medin, D. L. (1985). The role of theories in conceptual coherence. Psychological Review, 92, 289-316. Murphy, K. R. (1982). Difficulties in the statistical control of halo. Journal o f A pplied Psychology, 67, 161-164. Murphy, K. R. (1991). Criterion issues in performance appraisal research: Behavioral accuracy vs. classification accuracy. Organizational Behavior and Human Decision Processes, 50, 45-50. Murphy, K. R., & Balzer, W. K. (1986). Systematic distortions in memory-based behavior rat­ ings and performance evaluations: Consequences for accuracy. Journal o f A pplied Psychology, 71, 39-44. Murphy, K. R., & Balzer, W. K. (1989). Rater errors and rating accuracy. Journal o f Applied P sy­ chology, 74, 619-624. Murphy, K. R., Balzer, W. K., Lockhart, M. C., & Eisenman, E. J. (1985). Effects of previous performance on evaluations of present performance. Journal o f Applied Psychology, 70, 72-84. Murphy, K. R., Gannett, B. A., Herr, B. M ., & Chen, J. A. (1986). Effects of subsequent perform­ ance on evaluations of previous performance. Journal o f A pplied Psychology, 71, 427-431.

394

FELDMAN

Murphy, K. R., Garcia, M ., Kerkar, S., Martin, C ., & Balzer, W. K. (1982). Relationship between observational accuracy and accuracy in evaluating performance. Journal o f Applied P sychology, 68, 102-114. Murphy, K. R., Herr, B. M ., Lockhart, M ., & Maguire, E. (1986). Evaluating the performance of paper people. Journal o f Applied Psychology, 71, 654-661. Murphy, K. R., Philbin, T ., & Adams, S. (1989). Effect of purpose of observation on accuracy of immediate and delayed performance ratings. Organizational Behavior and Human Decision Process­ es, 43, 336-354. Nathan, B. R., & Alexander, R. A. (1985). The role of inferential accuracy in performance appraisal. Academy o f Management Review, 10, 109-115. Nathan, B. R., & Alexander, R. A. (1988). A comparison of criteria for test validation: A metaanalytic investigation. Personnel Psychology, 41, 517-535. Nathan, B. R., & Lord, R. (1983). Cognitive categorization and dimensional schemata: A process approach to the study of halo in performance ratings. Journal o f Applied Psychology, 68, 102-114. Neale, M. A., & Bazerman, M. H. (1991). Cognition and rationality in negotiation. New York: Free Press. Neisser, U. (1980). On “social knowing.” Personality and Social Psychology Bulletin, 60, 601-605. Newman, L. S., & Uleman, J. S. (1989). Spontaneous trait inference. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 155-188). New York: Guilford. Newtson, D. (1980). An interactionist perspective on social knowing. P ersonality and Social P sy­ chology Bulletin, 6, 520-531. Newtson, D., Engquist, G., & Bois, J. (1976). The reliability of a measure of behavior perception. JSAS Catalog o f Selected Documents in Psychology, 6, 5 (Ms. No. 1173). Olian, J., Schwab, D., & Haberfield, Y. (1988). The impact of applicant gender compared to qualifi­ cations on hiring recommendations. Organizational Behavior and Human Decision Processes, 41, 180-195. Oppler, S. H., Campbell, J. P., Pulakos, E. D., & Borman, W. C. (1992). Three approaches to the investigation of subgroup bias in performance measurement: Review, results and conclusions. Journal o f Applied Psychology (Monograph), 77, 201-217. Pazy, A. (1986). The persistence of pro-male bias despite identical information regarding causes of success. Organizational Behavior and Human Decision P rocesses, 30, 366-377. Pennington, N., & Hastie, R. (1992). Explaining the evidence: Tests of a story model for juror deci­ sion making. Journal o f Personality and Social Psychology, 62, 189-206. Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L. Ber­ kowitz (Ed.), Advances in experimental social psychology (Vol. 19, pp. 124-206). New York: Academic Press. Pulakos, E. D. (1984). A comparison of rater training programs: Error training and accuracy train­ ing. Journal o f Applied Psychology, 69, 581-588. Pulakos, E. D. (1986). The development of training programs to increase accuracy with different rating tasks. Organizational Behavior and Human Decision P rocesses, 38, 76-91. Pulakos, E. D., White, L. A., Oppler, S. H ., & Borman, W. C. (1989). Examination of race and sex effects on performance ratings. Journal o f Applied Psychology, 74, 770-780. Ragins, B. R., & Sundstrom, E. (1989). Gender and power in organizations: A longitudinal perspec­ tive. P sychological Bulletin, 105, 51-88. Read, S. J. (1987). Constructing causal scenarios: A knowledge structure approach to causal reason­ ing. Journal o f Personality and Social Psychology, 52, 288-302. Rosen, B., & Jerdee, T. H. (1976a). The nature of job-related age stereotypes. Journal o f Applied Psychology, 61, 180-183. Rosen, B., & Jerdee, T. H. (1976b). The influence of age stereotypes on managerial decisions. Jour­ nal o f A pplied Psychology, 61, 428-432.

7.

PERFORMANCE APPRAISAL

39 5

Rosen, B., Jerdee, T. H., & Lunn, R. O. (1981). Effects of performance appraisal format, age, and performance level on retirement decisions. Journal o f A pplied Psychology, 66, 515-519. Rosen, L ., & Fiske, S. T. (1990, August). Romantic outcome dependency and the (in)accuracy o f impression formation: A case o f clouded judgm ent. Poster presented at the meeting of the Ameri­ can Psychological Association, Boston, MA. Sackett, P. R., & Dubois, C. L. Z. (1991). Rater-ratee race effects on performance evaluation: Challenging meta-analytic conclusions. Journal o f Applied Psychology, 76, 873-877. Sackett, P. R., Dubois, C. L. Z ., & Noe, A. W. (1991). Tokenism in performance evaluation: The effects of work-group representation on male-female and Black-White differences in performance ratings. Journal o f A pplied Psychology, 76, 263-267. Salancik, G. R., & Pfeffer, J. (1977). An examination of need-satisfaction models of job attitudes. Administrative Science Quarterly, 22, 427-456. Schmidt, F. L., Hunter, J. E., & Outerbridge, A. N. (1986). Impact of job experience and ability on job knowledge, work sample performance, and supervisory ratings of job performance. Jour­ nal o f Applied Psychology, 71, 432-439. Schneer, J., & Reitman, F. (1990). Effects of employment gaps on the careers of MBAs: More damag­ ing for men than for women? Academy o f Management Journal, 33, 391-406. Schneider, B. J. (1983). Interactional psychology and organizational behavior. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 5, pp. 1-32). Greenwich, CT: JAI. Schneier, C. E. (1977). Operational utility and psychometric characteristics of behavioral expecta­ tion scales: A cognitive reinterpretation. Journal o f Applied Psychology, 62, 541-548. Schoorman, F. (1988). Escalation bias in performance appraisals: An unwanted consequence of su­ pervisors’ participation in hiring decisions. Journal o f Applied Psychology, 73, 58-62. Schwarz, N ., & Bless, H. (1991). Constructing reality and its alternatives: An inclusion/exclusion model of assimilation and contrast effects in social judgment. In L. Martin & A. Tesser (Eds.), The construction o f social judgments (pp. 217-245). Hillsdale, NJ: Lawrence Erlbaum Associates. Schweder, R. A. (1977). Likeness and likelihood in everyday thought: Magical thinking in judg­ ments of personality. Current Anthropology, 18, 633-648. Schweder, R. A., & D’Andrade, R. G. (1979). Accurate reflection or systematic distortion? A reply to Block, Weiss, and Thorne. Journal o f Personality and Social Psychology, 37, 1075-1084. Scott, W. A. (1969). Structure of natural cognitions. Journal o f Personality and Social Psychology, 12, 261-278. Sherman, S. J., Judd, C. M ., & Park, B. (1989). Social cognition. Annual Review o f Psychology, 40, 281-326. Showers, C., & Cantor, N. (1985). Social cognition: A look at motivated strategies. Annual Review o f Psychology, 36, 279-305. Sinclair, R. (1988). Mood, categorization breadth, and performance appraisal: The effects of order of information acquisition and affective state on halo, accuracy, information retrieval, and evalua­ tions. Organizational Behavior and Human Decision Processes, 42, 22-46. Slusher, M. P., & Anderson, C. A. (1987). When reality monitoring fails: The role of imagination in stereotype maintenance. Journal o f Personality and Social Psychology, 52, 653-662. Smith, D. E. (1986). Training programs for performance appraisal: A review. Academy o f Manage­ ment Review, 11, 22-40. Smith, P. C. (1976). Behavior, results and organizational effectiveness: The problem of criteria. In M. D. Dunnette (Ed.), Handbook o f industrial and organizational psychology (1st ed., pp. 745-776.) Chicago: Rand McNally. Smither, J. W ., Barry, S. R., & Reilly, R. R. (1989). An investigation of the validity of expert true score estimates in appraisal research. Journal o f Applied Psychology, 74, 143-151. Snyder, M. (1984). When belief creates reality. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 18, pp. 248-306.) New York: Academic Press. Snyder, M ., & Swann, W. B., Jr. (1978). Hypothesis-testing processes in social interaction. Journal o f Personality and Social Psychology, 36, 1202-1212.

396

FELDMAN

Srull, T. K., & Wyer, R. S., Jr. (1989). Person memory and judgment. Psychological R eview, 96, 58-84. Staw, B. M ., & Ross, J. (1985). Stability in the midst of change: A dispositional approach to job attitudes. Journal o f Applied Psychology, 70, 469-480. Stone, T. H., & Slusher, E. A. (1975). Attributional insights into performance appraisal. JSAS Catalog o f Selected Documents in Psychology, 5, 253 (Ms. No. 964). Sulsky, L. M ., & Balzer, W. K. (1988). The meaning and measurement of performance rating ac­ curacy: Some methodological and theoretical concerns. Journal o f Applied Psychology, 73, 497-506. Swim, J., Borgida, E., Maruyama, G., & Myers, D. (1989). Joan McKay vs. John McKay: Do gender stereotypes bias evaluations? Psychological Bulletin, 105, 409-429. Taylor, S. E., & Fiske, S. T. (1978). Salience, attention, and attribution: Top of the head phenome­ na. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 11, pp. 250-289.) New York: Academic Press. Tesser, A. (1978). Self-generated attitude change. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 11, pp. 290-338.) New York: Academic Press. Tetlock, P. E. (1985). Accountability: The neglected social context of judgment and choice. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 7, pp. 297-332). Greenwich, CT: JAI. Tetlock, P. E., & Boettger, R. (1989). Accountability: A social magnifier of the dilution effect. Journal o f Personality and Social Psychology, 57, 388-398. Thompson, D., & Thompson, T. (1985). Task-based performance appraisal for blue-collar jobs: Evalu­ ation of race and sex effects. Journal o f A pplied Psychology, 70, 747-753. Thornton, E. C., & Zorich, S. (1980). Training to improve observer accuracy. Journal o f Applied Psychology, 65, 351-354. Tourangeau, R., & Rasinski, K. A. (1988). Cognitive processes underlying context effects in atti­ tude measurement. Psychological Bulletin, 103, 299-314. Tsui, A., & Barry, B. (1986). Interpersonal affect and rating errors. Academy o f Management Jour­ nal, 29, 587-599. Tversky, A., & Kahneman, D. (1982). Evidential impact of base rates. In D. Kahneman, P. Solovic, & A. Tversky (Eds.), Judgment under uncertainty: Heuristics and biases (pp. 152-160). Cam­ bridge: Cambridge University Press. Waldman, D. A., & Avolio, B. J. (1986). A meta-analysis of age differences in job performance. Journal o f A pplied Psychology, 71, 33-38. Waldman, D. A., & Avolio, B. J. (1991). Race effects in performance evaluations: Controlling for ability, education and experience. Journal o f Applied Psychology, 76, 897-901. Weiner, B., Frieze, I., Kukla, A., Reed, L., Rest, S., & Rosenbaum, R. (1971). Perceiving the causes o f success and failu re . Morristown, NJ: General Learning Press. Wendelken, D. J., & Inn, A. (1981). Nonperformance influences on performance evaluations: A laboratory phenomenon? Journal o f Applied Psychology, 66, 149-158. Wexley, K. N., & Klimoski, R. (1984). Performance appraisal: An update. In K. R. Rowland & G. R. Ferris (Eds.), Research in personnel and human resources management (Vol. 2, pp. 35-79). Greenwich, CT: JAI. Wexley, K. N., & Pulakos, E. D. (1982). Sex effects on performance ratings in manager-subordinate dyads: A field study. Journal o f Applied Psychology, 67, 433-439. Wherry, R. J. (1952). The control o f bias in rating: A theory o f rating. (Personnel Research Board Rep. No. 922). Washington, DC: Department of the Army, Personnel Research Section. Williams, K. J. (1992). Introduction: The analytic and non-analytic frameworks for performance ap­ praisal. Human Resource Management Review, 2, 1-8. Williams, K. J., Cafferty, T. P., & DeNisi, A. S. (1990). The effect of performance appraisal salience on recall and ratings. Organizational Behavior and Human Decision Processes, 46, 217-239.

7.

PERFORMANCE APPRAISAL

397

Williams, K. J., DeNisi, A. S., Blencoe, A. G., & Cafferty, T. P. (1985). The role of appraisal purpose: Effects of purpose on information acquisition and utilization. Organizational Behavior and Human Decision P rocesses, 35, 314-339. Williams, K. J., DeNisi, A. S., Meglino, B. M ., & Cafferty, T. P. (1986). Initial decisions and subsequent performance evaluations. Journal o f Applied Psychology, 71, 189-195. Woehr, D. J., & Feldman, J. (1993). Processing objective and question order effects on the causal relation between memory and judgment in performance appraisal: The tip of the iceberg. Journal o f Applied Psychology, 78, 232-241. Woehr, D. J., & Huffcut, A. I. (1992). Rater training fo r performance appraisal. Manuscript sub­ mitted for publication. Wyer, R. S., & Budesheim, T. L. (1987). Person memory and judgment: The impact of information that one is told to disregard. Journal o f Personality and Social Psychology, 53, 14-29. Wyer, R. S., Jr., & Srull, T. K. (1980). Category accessibility: Some theoretical and empirical is­ sues concerning the processing of social stimulus information. In E. T. Higgins, C. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario symposium on personality and social psycholo­ gy (Vol. 1, pp. 161-198). Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., & Srull, T. K. (1986). Human cognition in its social context. Psychological Review, 93, 322-359. Wyer, R. S., & Unverzagt, W. (1985). Effects of instructions to disregard information on its subse­ quent recall and use in making judgments. Journal o f Personality and Social Psychology, 48, 533-549. Zajonc, R. B., (1960). The process of cognitive tuning in communication. Journal o f Abnormal and Social Psychology, 61, 159-167. Zalesny, M. D., & Highhouse, S. (1992). Accuracy in performance evaluations. Organizational Be­ havior and Human Decision Processes, 51, 22-50. Zedeck, S., & Cafry, D. (1977). Capturing rater policies for processing evaluation data. Organiza­ tional Behavior and Human Performance, 18, 269-294.

This page intentionally left blank

8

Consumer Judgment and Decision Processes

Frank R. Kardes University of Cincinnati

Contents Acquisition of Information About Products and Services 400 Selective Exposure 401 Selective Attention 402 Selective Encoding 408 Retention of Information About Products and Services 415 Brand Retrieval and Consideration Set Formation 415 Cognitive Organization of Product Information 417 The Relationship Between Memory and Judgment 420 Utilization of Information About Products and Services 423 Information Integration Theory 424 The Omission Detection Model 427 Data-Driven Versus Theory-Driven Inference Processes 429 The Availability-Valence Model 434 The Accessibility-Diagnosticity Model 438 The Hypothesis-Testing Model 442 Behavioral Decision Models 445 The MODE Model 448 Conclusion 450 References 451

Recent reviews have indicated that theory and research in consumer psychology continue to emphasize processes involved in judging, selecting, and using com­ mercial products and services (Bettman, 1986; Cohen & Chakravarti, 1990). Be­ cause of this process orientation, it is not surprising that consumer psychology is influenced heavily by theory and research in social cognition. However, con­ sumer psychology differs from social cognition in several respects: nonsocial stimuli such as products, services, advertisements, and other marketing-related stimuli serve as the targets of judgment; implications for marketing manage­ ment, business law, or public policy are emphasized; and consumer psychology is

399

40 0

KARDES

influenced heavily by theory and research in behavioral economics. Despite these points of departure, cognitive social psychologists are likely to find consumer research interesting, because the ubiquity and frequency of consumption-related thoughts and actions provide psychologists with the opportunity to investigate judgment, inference, and decision processes in an important everyday context. This chapter reviews recent research in consumer psychology bearing on the ac­ quisition, retention, and utilization of information concerning products and services. Many models of consumer information processing emphasize that, although a wide variety of products and services are commercially available, consumers tend to consider a relatively small set of alternatives when making purchase deci­ sions (Hauser & Wernerfelt, 1990; Kardes, Gurumurthy, Chandrashekaran, & Dornoff, 1993; Nedungadi, 1990). This finding is remarkably pervasive, despite that consumers are exposed to extensive amounts of information concerning a large number of different products and services every day. The typical American consumer is exposed to over 300 advertisements per day (Britt, Adams, & Miller, 1972). This estimate does not include exposure to product-related information provided by sales representatives, package labels, coupons, sweepstakes, pointof-purchase displays, consumer magazines, brochures, and friends and relatives. Due to capacity limitations (Miller, 1956), consumers process a relatively small proportion of the available information bearing on the products and services that are considered for purchase or use (Russo, 1977; Russo, Staelin, Nolan, Rus­ sell, & Metcalf, 1986). The information that is processed can have multiple im­ plications and can influence judgment in many different ways (Hoch & Deighton, 1989). These issues lead to a number of interesting questions: What information is considered when consumers judge products and services? How is product in­ formation stored and organized in memory? How is information used to make product-related judgments and decisions? ACQUISITION OF INFORMATION ABOUT PRODUCTS AND SERVICES The typical consumer is exposed to a relatively small subset of available infor­ mation about products and services, and the consumer attends to an even smaller subset of information to which he or she has been exposed. Not all information is encountered, and not all encountered information is attended to and processed. Because only encountered alternatives (i.e., brands, services, retail outlets) can be considered for brand choice, it is important to identify the antecedents of consumer-initiated external search for information concerning products and ser­ vices (e.g., retailer search and store visits; interactions with acquaintances, friends, and relatives; interactions with sales representatives; media search). Because some encountered alternatives will receive greater attention and consideration than others, it is important to understand selective attentional processes.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

401

Selective Exposure

How many alternatives and which alternatives are likely to be encountered by consumers? Most models of external search and selective exposure to product information emphasize that search effort is influenced by the perceived risk in­ volved in making a suboptimal decision (in terms of financial risk, performance risk, physical risk, social risk, and convenience risk) and the perceived benefits of search (Beatty & Smith, 1987; Duncan & Olshavsky, 1982; Furse, Punj, & Stewart, 1984; Kiel & Layton, 1981; Moore & Lehmann, 1980; Punj & Staelin, 1983; Srinivasan & Ratchford, 1991; Urbany, 1986; Urbany, Dickson, & Wilkie, 1989). The number of alternatives examined increases as perceived risk or as perceived benefits increase. Several variables influence perceived risk and per­ ceived benefits, including one’s prior knowledge and experiences pertaining to the target brand or product category, one’s interest in the target brand or product category, one’s attitudes toward shopping, and time availability. Decisions regarding frequently purchased packaged goods, such as margarine, coffee, toothpaste, cereal, snacks, laundry detergent, and so forth, are so incon­ sequential, and the benefits of search are so small that consumers engage in minimal search effort for these types of products (Hoyer, 1984; Hoyer & Brown, 1990; Olshavsky & Granbois, 1979). Perceived risk is low for inexpensive products that are replaced easily, and the benefits of search are low when the available alternatives are relatively homogenous. Consequently, consumers use very simple choice heuristics when purchasing mundane products. Heuristics such as the lexicographic heuristic (e.g., “Buy the least expensive brand”) or the familiar­ ity heuristic (e.g., “Buy the brand chosen last time” or “Buy the most familiar brand”) are used commonly. However, when shopping under time pressure in unfamiliar stores, consumers are less able to use simple choice heuristics, and, consequently, brand switching and unplanned purchasing increases (Park, Iyer, & Smith, 1989). Several studies have documented an inverted U-shaped relationship between level of prior knowledge and degree of search effort (Alba & Hutchinson, 1987; Bettman, 1979; Brucks, 1985; Johnson & Russo, 1984; Simonson, Huber, & Payne, 1988). Unknowledgeable consumers engage in little search effort, because they lack the mental representations and the cognitive resources needed to com­ prehend new information about unfamiliar products. Moderately knowledgeable consumers engage in extensive search effort, because they are able to compre­ hend new information, but they lack the knowledge and experience needed to ignore irrelevant information easily, to eliminate inappropriate alternatives quickly, or to make routine purchases. Highly knowledgeable consumers tend to engage in low levels of external search, because of their ability to rely almost exclusive­ ly on internal search (memory search). Moreover, highly knowledgeable con­ sumers may feel so confident about the scope and accuracy of their memories that they may feel little compulsion to seek additional information.

402

KARDES

Selective Attention

Consumers often lack the motivation, capacity, or opportunity to process all product information to which they are exposed in a thoughtful or deliberative manner. Goals influence the amount of attentional resources or cognitive effort consumers are willing to devote to the acquisition, interpretation, and elabora­ tion of product-related information (Batra & Ray, 1986; Cacioppo & Petty, 1982; Park & Young, 1986; Petty, Cacioppo, & Schumann, 1983; Yalch & ElmoreYalch, 1984). Arousal (Pavelchak, Antil, & Munch, 1988; Sanbonmatsu & Kardes, 1988) and prior knowledge (Alba & Hutchinson, 1987; Brucks, 1985; Johnson & Russo, 1984; Simonson et al., 1988) influence the ability to engage in effortful information processing. Situational factors such as time constraints, distractions, message repetition, message comprehensibility, and the manner and speed with which information is presented influence the opportunity to engage in effortful information processing (Cacioppo & Petty, 1985; Moore, Hausknecht, & Thamodaran, 1986; Petty, Wells, & Brock, 1976; Ratneshwar & Chaiken, 1991). Greenwald and Leavitt (1984) proposed a useful model of the relationship be­ tween consumer involvement (i.e., motivation/effort) and attentional processes. Motivation to engage in effortful information processing is viewed as a continu­ um ranging from very low to very high (Kahneman & Treisman, 1984; Shiffrin, 1988). When involvement is very low, information is processed with minimal effort, intention, or conscious control. Nevertheless, some information is likely to be encoded into long-term memory due to the operation of automatic, preattentive processes (“preattention”, Greenwald, 1988; Greenwald & Leavitt, 1984). When involvement is moderately low, focal attention is directed to goal-relevant, salient (contextually attention-drawing), or vivid (inherently attention-drawing) stimuli (Fiske & Taylor, 1991; Nisbett & Ross, 1980). Moderate involvement is required to relate new information to old information stored in long-term memory to facilitate comprehension and retention of new information (Alba & Hutchinson, 1987). Finally, when involvement is relatively high, consumers are likely to engage in elaborative processing, systematic thinking, and inferential reasoning that enables consumers to go “beyond the information given” (Bruner, 1957; Chaiken, 1980; Petty & Cacioppo, 1986). Preattention. Krugman (1965) suggested that consumers learn a great deal about products and services through television commercials that are processed incidentally with virtually no higher order cognitive activity. According to Krug­ man, the “public lets down its guard to the repetitive commercial use of the tele­ vision medium and that it easily changes its ways of perceiving products and brands and its purchasing behavior without thinking very much about it at the time of TV exposure or at any time prior to purchase, and without changing verbalized attitudes” (p. 354). Although Krugman never reported empirical tests of his ideas,

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

403

a few studies bearing on his propositions have appeared in the literature. For example, reserach on premature cognitive commitment has shown that little cog­ nitive effort is used to process goal-irrelevant information, and, as a consequence, irrelevant information often is accepted uncritically (Chanowitz & Langer, 1981). Later, if this information becomes relevant due to changing life circumstances, judgments and decisions tend to be influenced by beliefs that were formed previ­ ously on the basis of the uncritical acceptance of once irrelevant information. Similarly, consumers are unlikely to think extensively about unstimulating tel­ evision ads for mundane products. Why waste cognitive resources analyzing in­ formation and generating counterarguments regarding ads for unneeded things? Instead, much product information is likely to be processed incidentally and to be accepted at face value with little or no counterargumentation. However, goals change over time. Perhaps a currently used product will be consumed entirely, will break down, or will require replacement for any of a number of different reasons. Or perhaps a previously irrelevant product will become relevant because of new hobbies or interests. In any case, beliefs based on previously irrelevant information that was accepted uncritically may have an unwarranted influence on subsequent judgments and decisions. Recent research on preattentive processes has suggested that the subconcscious processing of nonfocal information and the organization of visual and verbal in­ formation in print ads influences evaluations of an advertised brand name (Janiszewski, 1988, 1990). Consistent with Friedman and Poison’s (1981) matching activation hypothesis, a brand is preferred more when presented to the right of pictorial information or to the left of textual information (Janiszewski, 1990). Moreover, the former effect is enhanced when subjects are instructed to attend to pictorial information, and the latter effect is increased when subjects are told to attend to verbal information. Additionally, evaluations of an advertised brand name increase when the amount of verbal information presented increases, provid­ ed that the brand name is presented to the left of the focal verbal material. When the brand name is presented to the right of this material, manipulations of the amount of verbal information presented has no effect on brand name evaluations. Together, these findings suggest that greater activation of the right (left) hemisphere during the processing of focal pictorial (verbal) information results in higher levels of subconscious processing of a nonfocal brand name performed by the other hemisphere. As processing increases, familiarity increases, and familiarity can be used as a cue to infer preference (Mandler, Nakamura, & Shebo Van Zandt, 1987). Focal Attention. Selective attention is directed or guided by intentional as well as unintentional cognitive processes (Kahneman, 1973). Moreover, focal attention can be directed outward toward product information present in the im­ mediate enviroment (external search) or inward toward product information stored in long-term memory (internal search). Information gleaned from the immediate

404

KARDES

environment or retrieved from long-term memory is stored temporarily in short­ term memory (conscious awareness). Because short-term memory can hold only seven plus or minus two chunks or units of information at a time (Miller, 1956; Newell & Simon, 1972), consumers attend to focal information and neglect nonfocal information. What factors determine whether a particular piece of product information is focal or nonfocal? What makes a particular piece of information stick out and grab people’s attention, whereas other pieces of information fade into the back­ ground? Momentary intentions and goals direct focal attention toward goal-related stimuli (Bruner, 1957; Kahneman, 1973). For example, food-related informa­ tion is attention drawing to someone who is hungry, beverage-related informa­ tion is attention drawing to someone who is thirsty, and entertainment-related information is attention drawing to someone who is bored (Klinger, 1975). Ad­ vertisements for lawnmowers, triple-pane windows, and automobiles are likely to capture the attention of consumers who need a new lawnmower, new win­ dows, or a new car, respectively. Klinger (1975) suggested that individuals’ “cur­ rent concerns” direct attention and color perception automatically. Although people have some voluntary control over attention (Kahneman, 1973), goals also appear to exert a nonvoluntary influence on selective attention (Erdelyi & Appelbaum, 1973; Klinger, 1975). Other nonvoluntary influences also guide focal attention. Some stimuli are too conspicuous to be ignored, even if people try to tune them out. Specifically, salient and vivid stimuli draw attention regardless of the consumer’s immediate inten­ tions or goals. Salient stimuli are attention drawing in some contexts, but not in others (context dependent), whereas vivid stimuli are attention drawing across contexts (context independent). For example, tall packages on the shelf in a grocery store are salient when the neighboring packages are short, but the same tall pack­ ages are nonsalient when the neighboring packages are tall. Similarly, brightly colored packages are salient when the neighboring packages are gray and drab, but the same brightly colored packages are nonsalient when the neighboring pack­ ages are brightly colored. By contrast, concrete, image-provoking stimuli are al­ ways more vivid than abstract, recondite stimuli. Not surprisingly, advertisers, package designers, and retailers are well aware of the importance of capturing the attention of consumers. Advertisers constant­ ly try to develop novel, creative, advertising executions, primarily because novel advertising executions are much more attention drawing than mundane, overused, or worn-out approaches. For example, the Energizer bunny television commer­ cials were very effective when they were first aired. The first Energizer bunny commercial began with an ad for a fictitious, humorous product that was unrelat­ ed to Energizer batteries. After a few seconds, the mock ad was interrupted by a battery-operated toy bunny marching across the television screen while bang­ ing a toy drum. Toward the conclusion of the ad, the narrator announced, “Ener­ gizer batteries. They keep going and going and going . . .” Initially, this ad was

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

405

very effective, and the advertiser was able to maintain a high level of advertising effectiveness for an extended period of time by developing numerous variations of the Energizer bunny theme (Schumann, Petty, & Clemons, 1990; Unnava & Burnkrant, 1991). Mock ads for many different fictitious products were devel­ oped, and each ad was interrupted by the marching and drumming toy bunny. Eventually, however, the novelty will wear off, consumers will tire of this exe­ cution, and the ad will no longer be effective (Calder & Sternthal, 1980). Hence, the pressure to develop novel, creative advertising executions never ends. Another type of novel advertising execution that has received recent empirical attention is the “mystery ad” (Fazio, Herr, & Powell, 1992). A mystery ad is a television commercial that withholds the identity of the advertised brand until the very end of the commercial. This approach is novel and attention drawing, and the results of this study indicated that, for unfamiliar brands, mystery ads were effective in building strong category-brand associations in memory. The authors suggested that withholding the identity of a brand arouses curiosity and creates a greater “readiness to categorize” the brand when its identity finally is revealed. A strong category-brand association results in the activation of the brand whenever the product category is brought to mind, and this increases the brand’s likelihood of being included in consumers’ consideration sets (Nedungadi, 1990). However, for familiar brands, mystery ads were no more effective than identical ads in which brands were identified early. Fazio et al. (1992) suggested that consumers may be disappointed when a mystery ad captures their interest and then provides an anticlimatic solution to the mystery (it’s just another ad for that old brand). Perceptual principles of Gestalt psychology suggest that figural stimuli are fo­ cal, whereas background stimuli are nonfocal. Bright, complex, and moving stimu­ li stand out when contrasted to drab, simple, and stationary backgrounds, respectively (Fiske & Taylor, 1991). Recent research on the judgmental effects of camera angles in print advertsing has suggested that a low camera angle pulls the advertised product into the foreground and pushes other stimuli presented in the ad into the background (Meyers-Levy & Peracchio, 1992). A low camera angle makes it seem like one is looking up at the product (an upward camera angle), whereas a high camera angle makes it seem like one is looking down on the product (a downward camera angle). An object may be more figural and salient when one looks up at it, relative to when one looks down on it or when one ex­ amines it at eye level (Kraft, 1987). Prior research has shown that salient objects are judged more extremely than nonsalient objects when personal involvement is low. However, when involvement is high, people are more likely to process information more thoroughly and systematically. Consequently, salience effects on judgment are reduced or eliminated when processing motivation is high (Borgida & Howard-Pitney, 1983; Howard-Pitney, Borgida, & Omoto, 1986). Con­ sistent with this research, Meyer s-Levy and Peracchio (1992) found that more favorable brand evaluations are formed when an upward camera angle is em­ ployed (compared with a downward or an eye-level camera angle), provided that

40 6

KARDES

processing motivation is low. When processing motivation is high, the camera angle effect on brand evaluations is eliminated. Prior knowledge also provides an important background or context for judg­ ing products and services. Unexpected information (i.e., information that is in­ consistent with prior knowledge and expectations) is very salient and attention drawing, and considerable cognitive effort is expended in attempts to resolve in­ consistencies between new versus old information. Consequently, unexpected in­ formation stimulates spontaneous causal reasoning (Hastie, 1984; Weiner, 1985), counterfactual reasoning (Gleicher et al., 1990; Kahneman & Miller, 1986; Meyers-Levy & Maheswaran, 1992), incongruity-based arousal and affect (Mandler, 1982; Meyers-Levy & Tybout, 1989; Stayman, Alden, & Smith, 1992), and spontaneous judgments of satisfaction or dissatisfaction (Oliver, 1980; Oliver & DeSarbo, 1988). Unexpected information can be unusual for a specific brand (e.g., new Coke is unusually sweet for a Coke) or for an entire product category (e.g., the Porsche 911 is unusually fast and sleek for an automobile). In both cases, consumer satis­ faction is based on product performance relative to initial expectations (Oliver, 1980; Oliver & DeSarbo, 1988; Stayman et al., 1992; Woodruff, Cadotte, & Jenkins, 1983). Consumer satisfaction increases when product performance ex­ ceeds consumers’ initial expectations. Conversely, dissatisfaction results when product performance falls below consumers’ initial expectations. Oliver and DeSar­ bo (1988) maintained that the degree of discrepancy between consumers’ initial expectations and subsequent consumption experiences is a more important deter­ minant of consumer satisfaction, relative to initial expectations or to product per­ formance per se. Satisfaction influences the frequency and direction of word-of-mouth communications (i.e., the extent to which consumers discuss their consumption experiences with friends and relatives, and the evaluative implica­ tions of these discussions) and the likelihood and duration of repeat purchase. Whereas the perceptual salience of a stimulus depends on the relationship be­ tween a stimulus and the context or background to which it is compared, the per­ ceptual vividness of a stimulus depends on the properties of the stimulus itself. Vivid stimuli are “(a) emotionally interesting, (b) concrete and imagery provok­ ing, and (c) proximate in a sensory, temporal, or spatial way” (Nisbett & Ross, 1980, p. 45). Of course, what is emotionally interesting to one consumer may be unintersting to another. Automobile-related stimuli (e.g., ads, pictures, arti­ cles, trade shows) are attention drawing to automobile enthusiasts, but not to con­ sumers who view automobiles simply as functional vehicles for getting from point A to point B. The product involvement scale provides a useful tool to assess the extent to which various product categories are emotionally interesting and vivid to a particular consumer (Zaichowsky, 1985). Some stimuli have general appeal (e.g., cute children, puppies, kittens, attractive models, humorous sketches) and are used to draw attention to an ad and to create a positive mood. Some of the positive affect generated by an emotionally interesting

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

407

ad generalizes to the advertised brand, even when the advertised brand and the emotionally interesting stimuli used in the ad are unrelated (Aaker, Stayman, & Hagerty, 1986; Brown & Stayman, 1992; MacKenzie, Lutz, & Belch, 1986; Mitchell & Olson, 1981; Stayman & Aaker, 1988). Recent research has suggest­ ed that the degree to which ad-based feelings and cognitions influence brand evalu­ ations depends on a number of moderating variables, such as processing goals, product familiarity, product category, advertising medium, advertising execu­ tion, advertising repetition, and characteristics of the consumer (Brown & Stay­ man, 1992; Miniard, Bhatla, & Rose, 1990; Olney, Holbrook, & Batra, 1991; Venkatraman, Marlino, Kardes, & Sklar, 1990). Concrete and imagery-provoking stimuli are also perceptually vivid. Product attributes that are described in concrete (vs. abstract) terms attract more atten­ tion and are weighed more heavily in judgment (Kisielius & Sternthal, 1984; MacKenzie, 1986), provided that the situation offers an opportunity for differen­ tial encoding and elaboration of vivid (vs. pallid) information (McGill & Anand, 1989; Taylor & Thompson, 1982). Concrete descriptions provide detailed and specific information about product performance that make it easy to imagine pre­ cisely how the product operates. Abstract descriptions are general, nonspecific, and difficult to visualize. Unfortunately, one ambiguity raised by the con­ crete/abstract manipulation of information vividness is that information content also is varied (Kisielius & Sternthal, 1986; Taylor & Thompson, 1982). One way to avoid this ambiguity is to hold information content constant and to manipulate the manner in which information is presented. Herr, Kardes, and Kim (1991) provided small groups of four to six subjects with attribute informa­ tion “condensed from Consumer Reports,” pertaining to a new personal computer. In addition to this attribute information, subjects received positive or negative anecdotal information that was presented either in a vivid, face-to-face format or in a pallid, printed format. In vivid format conditions, the anecdotal informa­ tion was presented in a word-of-mouth communication delivered by a confeder­ ate. In pallid format conditions, the anecdotal information was presented in a printed paragraph transcribed from an interview with a college student. The anec­ dotal information was held constant, word for word, regardless of format. Moreover, the anecdotal information and the attribute information always had contradictory evaluative implications. In vivid anecdote conditions, product evalu­ ations were influenced more heavily by anecdotal than by attribute information. By contrast, in pallid anecdote conditions, product evaluations were influenced more heavily by attribute than by anecdotal information. A second experiment revealed that the impact of vivid anecdotal information on product judgments was reduced when a previously formed impression of the described brand was availa­ ble from memory or when extremely negative attribute information was provided. Several studies have suggested that pictorial (vs. verbal) information is more vivid, attention drawing, memorable, and impactful on judgment (the picturesuperiority effect; Childers & Houston, 1984; Houston, Childers, & Heckler,

408

KARDES

1987; Mitchell, 1986). When consumers were instructed to focus on the physical appearance and layout of ads, brand name recall was greater for pictorial (vs. verbal-only) ads in both immediate- and delayed-recall conditions (Childers & Houston, 1984). When consumers were instructed to focus on the semantic con­ tent of ads, pictorial and verbal-only ads were equally effective in immediate recall conditions. However, after a delay, the picture-superiority effect reemerged. The picture-superiority effect was enhanced when pictorial and verbal information had inconsistent judgmental implications (Houston et al., 1987) or when consumers were primarily visually oriented (vs. verbally oriented) information processors (Childers, Houston, & Heckler, 1985). Finally, a proximal stimulus is more vivid than a distal stimulus. Firsthand, recently encountered, and local information is more vivid than secondhand, histor­ ical, and nonlocal information, respectively (Nisbett & Ross, 1980). Events relayed by another individual, that took place long ago, or that occurred in some faraway land are not highly attention drawing and are not weighed heavily in judgment. Information that is obtained through direct firsthand experience (Fazio & Zanna, 1981; Smith & Swinyard, 1983), that is obtained recently (Schwartz, 1978), or that has clear and direct implications for personal outcomes (Sivacek & Crano, 1982) is more attention drawing and impactful on judgment. Selective Encoding

Just as people focus on some aspects, but neglect other aspects of externally presented stimuli (selective attention), so too do people focus on some aspects, but neglect other aspects of stimuli held in short-term memory (selective encod­ ing). Attention refers to the contents of short-term memory, and these contents can be drawn from externally presented stimuli, from internal representations stored in long-term memory, or from both external and internal sources (Kahneman, 1973). Encoding refers to the process by which external stimuli are trans­ formed into internal cognitive representations. Processing goals, individual difference variables, and contextual variables influence this process. Comprehension. Comprehension involves the formation of interpretive in­ ferences pertaining to the intended meaning of a statement (Alba & Hutchinson, 1987). Interpretive inferences are important, because they facilitate communica­ tion by enabling consumers to glean the meaning of a product claim without sys­ tematically considering and evaluating each possible meaning of each word presented in a message. Examples of interpretive inferences include coherence inferences (i.e., inferences concerning relationships between presented statements; Keenan, Baillet, & Brown, 1984), instrumental inferences (e.g., the phrase “he pounded a nail into the wall” implies a hammer was used; Corbett & Dosher, 1978), and pragmatic implications (i.e., assumptions and conclusions based on statements that are literally, but not figuratively true; Harris, 1977; Harris &

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

409

Monaco, 1978). Generally, interpretive inferences are formed readily with rela­ tively little effort. In fact, so little effort typically is involved in the generation of interpretive inferences that consumers frequently experience difficulty in dis­ criminating between presented information and interpretive inferences formed on the basis of this information (e.g., did the text literally state that he used a hammer, or did I infer this? Alba, 1984; Arkes & Freedman, 1984; Kardes, 1988a). Hasher, Goldstein, and Toppino (1977) examined the effects of repetition on the perceived validity of statements pertaining to a wide variety of general knowledge topics (e.g., science, politics, art, history). Although no feedback was provided bearing on the objective validity of a statement, the perceived validity of a statement increased with repeated presentations of the statement over a 4-week period. Subsequent research has shown that the repetition-induced increase in judged validity (or the truth effect) does not depend on the objective validity of a statement. Moreover, the truth effect is more pronounced when subjects are somewhat familiar (vs. unfamiliar) with the topic of a statement (Arkes, Boehm, & Xu, 1991; Arkes, Hackett, & Boehm, 1989). When people are unable or unwilling to judge the validity of a statement direct­ ly, they may infer validity on the basis of cues available in a memory trace (Roediger, 1990). If a statement seems familiar and appears to “ring a bell,” peo­ ple assume that the statement is probably true, even if they cannot remember the source of the information. Familiarity may serve as a heuristic cue to validi­ ty, just as availability serves as a heuristic cue to probability (Tversky & Kahne­ man, 1974). When heuristic thinking is likely (i.e., when the motivation, ability, or opportunity to process information systematically is low), repetition increases familiarity and familiarity may increase perceived validity. If mere repetition is sufficient to increase the perceived validity of statements, the truth effect has straightforward implications with respect to the effects of repeti­ tive advertising on the perceived validity of product claims. To investigate the extent to which the truth effect generalizes to product claims and to examine the presumed mediating role of familiarity, Hawkins and Hoch (1992) exposed sub­ jects to a large set of product claims and manipulated subjects’ initial processing goals. The product claims pertained to a wide variety of products and services (e.g., “Consumers are more satisfied with State Farm homeowner’s insurance than with Allstate,” “Prolonged use of Alka Seltzer can lead to the formation of kid­ ney stones,” “Antehistamines have no effect on the common cold,” “Stone-ground flour retains more nutrients than conventional flour”). Initial processing goals were manipulated by instructing subjects to judge the validity of each statement each time it appeared on the computer monitor (high involvement) or to judge the com­ prehensibility of each statement (low involvement). Repetition (one or two ex­ posures), exposure time (3 vs. 8 seconds), and delay between Sessions 1 and 2 (a 20-minute vs. a 1-week delay) also were manipulated. As expected, judged validity was greater for repeated than for nonrepeated

410

KARDES

product claims. Moreover, this effect was more pronounced when subjects were instructed to judge the comprehensibility (vs. the validity) of the claims during the acquisition phase of the experiment. Path analyses revealed that familiarity (recognition accuracy) increased with repetition, and judged validity increased with familiarity. Although repetition had a stronger impact on familiarity in evalu­ ation (high-involvement) than in comprehension (low-involvement) conditions, familiarity played a less important mediating role in high- than in low-involvement conditions. This pattern suggests that subjects may be more sensitive to and more likely to make allowances for the biasing effects of familiarity on perceived va­ lidity when initial processing effort his high, perhaps because they may be more likely to recognize that repetition influenced familiarity when initial processing effort is high. However, subjects may be less likely to recognize and make al­ lowances for the biasing effects of familiarity on judged validity when initial processing effort is low. Similarly, consumers are most susceptible to misleading advertisements present­ ing claims that are literally true, but figuratively false, when processing effort is relatively low (Harris, 1977; Harris & Monaco, 1978). For example, the claim “Brand X may help you lose weight” is literally true (the product may or may not help you lose weight), but figuratively false (in everyday language, may im­ plies usually). Similarly, the claim that “Brand X gives you better gas mileage” is literally true (the product gives you better gas mileage than water), but figura­ tively false (consumers are likely to assume that Brand X gives you better gas mileage than other brands of gasoline). Consumers are also likely to infer that the claim “Women who look younger use Oil of Olay” indicates that “women who use Oil of Olay look younger.” “Confusion of the inverse” is a logical fallacy that is committed frequently in everyday reasoning (Dawes, 1988; Gavanski & Hui, 1992; Klar, 1990; Sherman, McMullen, & Gavanski, 1992; Wyer, 1977; Wyer & Srull, 1989). Wyer and Srull (1989) suggested that confusion of the inverse underlies many different forms of heuristic thinking. Specifically, “subjects who infer that a stimu­ lus condition implies another (that X implies Y) will also believe that the second implies the first (that Y implies X) and, therefore, will infer X from the existence of Y” (Wyer & Srull, 1989, p. 281). This general inference principle can be ap­ plied to a wide range of cognitive heuristics (for reviews of cognitive heuristics used to simplify judgment and decision making, see Nisbett & Ross, 1980; Sher­ man & Corty, 1984; Tversky & Kahneman, 1974). For example, the represent­ ativeness heuristic involves using similarity as a cue forjudging probability. Many consumers believe that if a brand is a member of a certain product category, it is likely to possess features similar to the features of the product category. In­ verting X and Y results in the potentially misleading belief that if a brand pos­ sesses features similar to the features of a product category, the brand is likely to be a member of that category. This belief is potentially misleading, because probability of category membership is influenced by many factors other than per­

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

411

ceived similarity (e.g., statistical considerations). The availability heuristic in­ volves using the ease with which examples are retrieved from memory as a cue forjudging probability. Most consumers believe that frequently occurring events are memorable. Inverting X and Y results in the belief that memorable events are frequently occurring. This belief is potentially misleading, because memory is influenced by many factors other than objective frequency (e.g., salience, vivid­ ness, memory structure and organization). The simulation heuristic involves using the ease with which events are imagined as a cue for judging probability. Many people believe that likely events are easy to imagine. Inverting X and Y results in the belief that events that are easy to imagine are likely to occur. This belief is potentially misleading, because imagination is influenced by many factors other than objective frequency (e.g., memory structure and organization, wishful think­ ing). Confusion of the inverse also can be used to explain salience and vividness effects on judgment. Consumers believe that important information is attention drawing. Inverting X and Y results in the belief that attention drawing informa­ tion is important. This belief is potentially misleading, because information im­ portance or relevance in influenced by many factors other than attention (e.g., statistical considerations, diagnosticity). Other examples of confusion of the in­ verse and heuristic thinking were provided by Wyer and Srull (1989). Elaboration. Fortunately, consumers do not always engage in heuristic think­ ing. Important decisions are likely to be guided by judgments formed on the ba­ sis of a systematic analysis of a relatively comprehensive set of judgment-relevant evidence, provided that the decisionmaker possesses the ability and the opportu­ nity to engage in effortful information processing. The Elaboration Likelihood Model (ELM; Petty & Cacioppo, 1986; Petty et al., 1983; Petty, Unnava, & Strathman, 1991; Schumann et al., 1990) and the Heuristic-Systematic Model (HSM; Chaiken, 1980; Chaiken, Liberman, & Eagly, 1989; Maheswaran, Mackie, & Chaiken, 1992; Ratneshwar & Chaiken, 1991) suggested that there are two qualitatively different paths or routes to persuasion. The central (or systematic) route involves the careful analysis and integration of all available judgment-relevant evidence when evaluating a given attitudinal position. When following the cen­ tral route to persuasion, consumers are unlikely to process information about products and services passively. Instead, consumers are likely to elaborate on a persuasive message by generating their own arguments or cognitive responses (i.e., support arguments, counterarguments, source derogations, and curiosity statements; Batra & Ray, 1986; Greenwald, 1968; Hastak & Olson, 1989; Petty, Ostrom, & Brock, 1981; Wright, 1980). Consumers often find their own self­ generated arguments to be more compelling than the original message arguments that provided the impetus for cognitive elaboration (Greenwald, 1968; Petty et al., 1981). Attitudinal judgments formed on the basis of highly compelling argu­ ments are likely to be relatively enduring, resistant to change, and likely to in­ fluence subsequent behavior.

412

KARDES

As emphasized previously, consumers are not always willing and able to en­ gage in effortful information processing. The resolution of many everyday problems and decisions does not require the allocation of a large proportion of an individual’s limited cognitive capacity. When the motivation, ability, or op­ portunity to engage in effortful information processing is low, elaboration likeli­ hood is also low. Motivation is low when involvement is low (Batra & Ray, 1986; Park & Young, 1986; Petty et al., 1983; Yalch & Elmore-Yalch, 1984) or when need for cognition is low (Cacioppo & Petty, 1982). Ability is low when arousal is high (Pavelchak et al., 1988; Sanbonmatsu & Kardes, 1988) or when domainspecific prior knowledge is low (Alba & Hutchinson, 1987; Brucks, 1985; John­ son & Russo, 1984; Simonson et al., 1988). Opportunity is low when distraction is high (Petty et al., 1976), message repetition is low (Cacioppo & Petty, 1985), message comprehensibility is low (Ratneshwar & Chaiken, 1991), or when a mes­ sage is presented at a very fast pace, as in time-compressed television advertis­ ing (Moore et al., 1986). When elaboration likelihood is low, consumers are likely to follow the peripheral route to persuasion. Instead of carefully evaluating the quality of the arguments supporting or refut­ ing a given attitudinal position, relatively superficial cues are used to form an attitude quickly and easily when consumers follow the peripheral route to persua­ sion. Many different types of superficial cues may be used. For example, the HSM suggests that several heuristics forjudging the validity of an attitudinal po­ sition are stored in long-term memory. The “length implies strength” heuristic implies that attitudinal positions supported by a large number of arguments are likely to be valid. The “experts are generally correct” heuristic implies that at­ titudinal positions endorsed by experts are likely to be valid. Of course, argu­ ment quality tends to be neglected when consumers rely too heavily on simplifying heuristics. In addition to the HSM persuasion heuristics, the ELM suggests that other types of simple cues may be used when consumers follow the peripheral route to persuasion. Pleasant music, attractive models, cute babies, and other stimuli that elicit a positive emotional response are used in advertising to influence brand attitudes through classical conditioning (Allen & Janiszewski, 1989; Allen & Madden, 1985; Bierley, McSweeney, & Vannieuwkerk, 1985; Gorn, 1982; McSweeney & Bierley, 1984; Shimp, Stuart, & Engle, 1991; Stuart, Shimp, & Engle, 1987). Category-based inferences also provide a quick and easy way to form atti­ tudes and opinions (Fiske & Neuberg, 1990; Fiske & Pavelchak, 1986; Sujan, 1985). Consumers may infer that if a product is made in Japan, it is probably of high quality. Or if a product is carried by K-Mart, it is probably cheap. Category-based inferences enable consumers to form attitudes with virtually no information. A consumer who likes Indian cuisine may infer that he or she prob­ ably would like tandoori chicken even if he or she had never tried or seen it. Attitudes also can be inferred on the basis of superficial behavioral information. For example, Bern’s (1972) self-perception theory suggested that a consumer may

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

4 13

infer that he or she likes brown bread, simply because he or she is eating brown bread. Very little analysis or effort is required to draw such a simple attitudinal inference. Many different types of peripheral, superficial cues can serve as a ba­ sis for attitude formation or change when consumers are unwilling or unable to engage in effortful information processing. The ELM suggests that attitudes formed through the peripheral (vs. central) route are less persistent, resistant, and predic­ tive of behavior. Nevertheless, attitudes formed on the basis of simple peripheral cues have the capacity to influence subsequent purchase decisions (Miniard, Sirdeshmukh, & Innis, 1992). Although the ELM has proved to be an extremely useful framework for analyz­ ing advertising effectiveness, several key issues remain to be resolved by future research. Perhaps the most controversial issue pertains to the nature and struc­ ture of the involvement construct. The ELM suggests that involvement is a general motivational variable that induces effortful information processing across a wide variety of situations (Petty & Cacioppo, 1990). However, several researchers have suggested that there are qualitatively different forms of involvement, and different motives can induce different processes (Chaiken & Stangor, 1987; John­ son & Eagly, 1989, 1990; Kruglanski, 1989). For example, Kruglanski’s (1989) theory of lay epistemology differentiated between three different types of infor­ mation processing objectives: need for structure, fear of invalidity, and wishful thinking. Need for structure (or need for closure) refers to a preference for knowledge, any knowledge, as opposed to ambiguity and confusion. This goal is likely to be salient when consumers need to make a purchase decision under time pressure, and this goal motivates consumers to form a belief toward an ob­ ject or issue relatively quickly (beliefs are the building blocks of knowledge). Fear of invalidity (need to avoid closure) refers to the concern about forming an erroneous belief that could lead to aversive consequences. This goal is likely to be salient when consumers need to make a purchase decision concerning a product that has to operate smoothly for many years (e.g., an automobile, a kitchen appliance, a television), and this goal motivates consumers to think carefully about an issue and to withhold judgment until sufficient evidence to generate an informed opinion has been accumulated. Need for structure promotes epistemic freezing (i.e., closed-mindedness), whereas fear of invalidity promotes epistemic unfreezing (i.e., open-mindedness). Finally, wishful thinking refers to a preference for be­ liefs implying that pleasant outcomes are likely, whereas aversive outcomes are unlikely (e.g., postdecisional dissonance reduction). Interestingly, Kruglanski’s (1989) information-processing goals seemed some­ what related to the attitude functions proposed by Katz (1960) and Smith, Bruner, and White (1956). Need for structure, fear of invalidity, and wishful thinking map on to the knowledge function (attitudes can provide convenient summaries of large amounts of information), the adjustment function (attitudes can facilitate the avoidance of aversive consequences), and the ego-defensive function (atti­ tudes can serve a self-esteem maintenance function). One additional attitude

414

KARDES

function was proposed by Katz (1960) and Smith et al. (1956): the value-expressive function. This function emphasized that attitudes can be used as tools to control or manage the impressions others form of people during interpersonal communi­ cation. Hence, research on attitude functions suggests that there may be four differ­ ent types of involvement: issue involvement, outcome involvement, ego-defensive involvement, and identity involvement. Another important topic for future research is Kruglanski’s (1989) hypothesis that attitudes formed on the basis of peripheral cues can be as persistent, re­ sistant to persuasion, and impactful on behavior as attitudes formed on the basis of highly compelling arguments. For example, an expert source is a pe­ ripheral cue that can be processed using the heuristic that “experts are usually correct.” Kruglanski (1989) emphasized that many religious leaders can induce their followers to form strongly held beliefs in the absence of carefully reasoned arguments and logically compelling evidence. Similarly, an attractive source is a peripheral cue that can be processed using the heuristic that “attractive peo­ ple are trustworthy.” However, Kahle and Homer (1985) demonstrated that an attractive source is not always a peripheral cue. Advertisements employing physically attractive endorsers of health and beauty aid products, fashion products, and fashion accessories can be very effective, because the attractive sources themselves provide a direct statement about the benefits of using these types of products. Whereas Kahle and Homer’s (1985) results suggest that strong attitudes can be formed on the basis of peripheral cues, Kardes, Kim, and Lim’s (in press) data suggest that weak attitudes can be formed on the basis of central cues (both effects were anticipated by Kruglanski’s theory of lay epistemics). Kardes, Kim, and Lim (in press) exposed consumers to product claims implying specific con­ clusions regarding the performance of an advertised brand. These conclusions either were stated directly (explicit arguments) or were omitted (implicit argu­ ments) from the text of a mock ad. Subjects also were blocked into high versus low prior knowledge conditions on the basis of their scores on a product knowl­ edge survey. Although high- and low-knowledge subjects formed identical be­ liefs regarding product performance, only low-knowledge subjects questioned the validity of their self-generated beliefs. Moreover, low-knowledge subjects rated explicit arguments as more diagnostic than self-generated arguments. By contrast, high-knowledge subjects tended to rate self-generated arguments as more diagnostic than explicit arguments. Hence, contrary to the basic premise of the cognitive response approach to persuasion (Greenwald, 1968; Petty et al., 1981), self-generated arguments are not always more compelling than arguments provided by others. Belief and attitude strength may be influenced more by hypothesis (judgment) formation and validation processes (Kruglanski, 1989) than by the type of evidence (i.e., central vs. peripheral cues) serving as a basis for judgment.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

415

RETENTION OF INFORMATION ABOUT PRODUCTS AND SERVICES

Because there is often a considerable gap between the point in time at which in­ formation about products and services is acquired and the point in time at which information about products and services is used to make purchase decisions, it is of crucial importance to develop an understanding of the role of memory in consumer judgment and decision making (Alba, Hutchinson, & Lynch, 1991; Baker & Lutz, 1988; Lynch & Srull, 1982). Alba et al. (1991) maintained that long-term memory can influence several dimensions of brand choice, including what brands are considered, what information is used to evaluate considered brands, and how this information is used. Lynch and Srull (1982) distinguished between three general types of brand choice situations: stimulus-based choice, memory-based choice, and mixed-choice. Stimulus-based choice involves the selection of one brand from a set of brands that are physically present at the time of choice. Occasionally, grocery shoppers consider only brands on shelves directly in front of them, but Alba et al. (1991) suggested that purely stimulus-based de­ cisions are made relatively infrequently. Memory-based choice involves select­ ing one brand from a set of brands stored in memory with none of the alternatives being physically present at the time of choice. Choosing a restaurant before one leaves one’s home or office is a good example of a purely memory-based choice. Mixed choice involves considering alternatives retrieved from memory along with alternatives that are physically present at the time of choice. This type of choice decision occurs when consumers compare products from different stores that carry different brands or models. Brand Retrieval and Consideration Set Formation

For most product categories, a large number of alternative brands are available in today’s cluttered marketplace. If one glances through any recent issue of Con­ sumer Reports, it becomes immediately apparent that there are at least 15 differ­ ent brands and models to choose from in most product categories. However, the actual number of brands available in the marketplace (the universal set) is much larger, because Consumer Reports evaluates a limited set of alternatives. Neverthe­ less, for most product categories, consumers are aware of a small set of alterna­ tives (the awareness set), and the set of brands that actually are considered for purchase (the consideration set) is smaller still. For many product categories, the average number of brands included in consumers’ consideration sets ranges from two to eight (Hauser & Wernerfelt, 1990). As the number of brands includ­ ed in consumers’ consideration sets increases, the average probability of choos­ ing any given brand decreases (e.g., p = 1.00, 0.50, 0.33, 0.25, and 0.20 for con­ sideration set sizes of 1, 2, 3, 4, and 5, respectively). Moreover, as consideration

416

KARDES

set size increases, the likelihood of retrieving additional brands from memory decreases due to part-list cuing (Alba & Chattopadhyay, 1985, 1986; Rundus, 1973). What factors influence the likelihood with which a particular brand will be included in consumers’ consideration sets? Prototypical brands, or brands that resemble most other members of a product category, are more accessible from memory than less prototypical brands (Loken & Ward, 1990; Nedungadi & Hutch­ inson, 1985; Rosch & Mervis, 1975). Brands that are associated strongly with a particular goal or usage situation (e.g., “Weekends are made for Michelob”) are primed or activated from memory whenever that particular goal or usage sit­ uation arises (Barsalou, 1985; Barsalou & Ross, 1986; Ratneshwar & Shocker, 1991). Brands that are evaluated very positively or very negatively are more ac­ cessible from memory, relative to moderately evaluated brands (Judd & Kulik, 1980; Pratkanis & Greenwald, 1989). Frequently and recently activated brands are also accessible from memory (Anderson, 1983; Higgins, Bargh, & Lombardi, 1985; Nedungadi, 1990; Wyer & Srull, 1986, 1989). Elaborative encoding can enhance brand accessibility by increasing the number of cues in a memory trace that can be used subsequently to facilitate brand retrieval (Anderson, 1983; Nedun­ gadi, 1990; Wyer & Srull, 1986, 1989; but, for an alternative view, see Walker, 1986). Recent research has shown that the first brand to enter a new product market (the pioneering brand) often is preferred to later entrants or follower brands (Car­ penter & Nakamoto, 1989; Gurumurthy & Urban, 1992; Kardes & Gurumurthy, 1992; Robinson & Fornell, 1985; Urban, Carter, Gaskin, & Mucha, 1986). This pioneering advantage is at least partially due to the readiness or ease with which the pioneering brand comes to mind when consumers think about a product category (Kardes, Gurumurthy, Chandrashekaran, & Dornoff, 1993). The pi­ oneering brand is more accessible than follower brands for many reasons: the pioneering brand is more prototypical (Carpenter & Nakamoto, 1989), more fre­ quently activated (i.e., consumers often are exposed to a large amount of infor­ mation about the pioneering brand before the second entrant appears on the market, thus giving the pioneering brand a headstart advantage; Kardes & Gurumurthy, 1992), and more distinctive and attention drawing (Kardes, Gurumurthy, Chan­ drashekaran, & Dornoff, 1993). Initially, the pioneering brand is the only brand that can be retrieved and included in consumers’ consideration sets (at least until the appearance of the second entrant). Pioneering can influence brand consideration indirectly through its influence on brand retrieval. Brands that are retrieved readily are likely to be included in con­ sumers’ consideration sets (Nedungadi, 1990). Additionally, pioneering can in­ fluence brand consideration directly, because, for the pioneering brand, considera­ tion benefits are likely to exceed consideration costs. Hauser and Wernerfelt’s (1990) evaluation cost model suggests that the costs involves in deciding whether or not to include a given brand in the consideration set—including thinking costs

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

417

(Alba & Hutchinson, 1987; Shugan, 1980), search costs (Beatty & Smith, 1987), and opportunity costs (Schmalensee, 1982)—are low for the pioneering brand, initially, because it is the only brand available for consideration. Information load and evaluation costs increase as more brands enter the market. Evaluation benefits are high for the pioneering brand, initially, because it is the only relevant brand. Evaluation benefits increase with consideration set size, because the optimal brand (i.e., the brand with the greatest expected utility) is more likely to be considered as consideration set size increases. However, the marginal incremental utility of adding an additional brand to the consideration set decreases as consideration set size increases (because the optimal brand already may have been included in the consideration set). Kardes, Gurumurthy, Chandrashekaran, and Dornoff (1993) provided evidence for indirect and direct effects of pioneering on consideration set composition. Pioneering influences brand consideration indirectly by first af­ fecting brand retrieval. Moreover, the pioneering brand is likely to be considered even when variability due to brand retrieval is controlled. The pioneering brand is also likely to be chosen, even when the indirect effects of brand retrieval and brand consideration on choice are controlled. Cognitive Organization of Product Information

Early research on the cognitive representation of product information focused on the effects of information processing goals and information presentation for­ mat on memory structure (Bettman, 1979; Bettman & Park, 1980; Biehal & Chakravarti, 1982, 1983). Processing goals were manipulated through experimen­ tal instructions and presentation format was controlled through the use of infor­ mation display boards (Payne, 1976). Protocol analysis (Ericsson & Simon, 1980) was used to determine memory organization. A brand-based memory organiza­ tion is characterized by a separate location (or storage bin; Wyer & Srull, 1986, 1989) for each brand stored in long-term memory. Each target brand bin con­ tains a list of attributes possessed by the target brand. By contrast, an attributebased memory organization is characterized by a separate bin for each attribute stored in long-term memory. Each target attribute bin contains a list of brands possessing the target attribute. When subjects are instructed to memorize product information, this informa­ tion tends to be organized in memory by brand (Biehal & Chakravarti, 1982, 1983). Brand-based memory structures are also likely when product information is presented in a brand-based format (i.e., each brand is described separately). Conversely, when subjects are instructed to choose the best brand, or when in­ formation is presented in an attribute-based format (i.e., each attribute is presented separately, along with the brands possessing the attribute), greater attribute-based memory organization follows. The manner in which product information is stored in memory then influences comparison processes and brand choice (Biehal & Chakravarti, 1982, 1983). A brand-based memory organization facilitates brand-

418

KARDES

based comparison processes, whereas an attribute-based memory organization facilitates attribute-by-attribute comparisons across brands. However, when a large set of alternatives are available during stimulus-based choice, consumers simplify choice by engaging in a two-step phased decision process. First, non­ compensatory choice heuristics (e.g., conjunctive, disjunctive, lexicographic, elimination-by-aspects heuristics; Bettman, 1979; Bettman, Johnson, & Payne, 1991) are applied to reduce the number of brands considered. Second, this reduced set of brands is subjected to a compensatory evaluation process. More than two steps or phases are likely during complex mixed-choice tasks (Biehal & Chakravarti, 1986). The idea that information is linked together in an associative network is one of the most widely accepted principles of memory organization (Anderson, 1983; Anderson & Bower, 1973; Carlston & Skowronski, 1986; Raaijmakers & Shiffrin, 1981; Wyer & Carlston, 1979). Evidence for associative memory stems from two general sets of findings: (a) when two cues are linked together in memory, activation of one cue facilitates activation of the other (the priming effect; for examples of priming in consumer contexts, see Bettman & Sujan, 1987; Herr, 1989; Keller, 1987; Nedungadi, 1990; Yi, 1990); and (b) as the number of cues linked together in memory increases, likelihood of activation of a particular cue decreases (associative interference; for examples of associative interference in consumer contexts, see Burke & Srull, 1988; Keller, 1987, 1991; Tybout, Calder, & Sternthal, 1981). An association between two cues is formed simply by thinking about the cues simultaneously, and the strength of an association increases with frequent or re­ cent activation (Anderson, 1983). Effortful encoding also can increase the strength of an association (Tyler, Hertel, McCallum, & Ellis, 1979). Anderson (1983) suggested that elaborative processing can facilitate recall directly by providing multiple pathways for accessing a particular cue or indirectly by providing mul­ tiple cues to facilitate inferential reconstruction of portions of a memory trace that are difficult to retrieve directly. Walker (1986) found that elaborative process­ ing inhibits direct retrieval, but facilitates inferential reconstruction. Many memory researchers have suggested that activation spreads automatically across an associa­ tive network (Anderson, 1983; Collins & Loftus, 1975), but Ratcliff (1981; see, also, Ratcliff & McKoon, 1988) found that facilitation begins at about the same time, regardless of the number of nodes between a priming cue and a target cue. However, the question regarding whether activation spreads automatically is not near resolution (McNamara, 1992). Herr (1989) unobtrusively primed different price categories (i.e., very expen­ sive, moderately expensive, moderately inexpensive, very inexpensive) by ex­ posing subjects to brands representative of these categories. When subjects were primed with moderately expensive or moderately inexpensive brands, subsequent judgments of an unfamiliar target brand shifted toward the primed category (as­ similation effects were observed). That is, the target brand was rated as more

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

4 19

(less) expensive when subjects were exposed previously to moderately expensive (inexpensive) brands. Conversely, when subjects were primed with extremely expensive or extremely inexpensive exemplars, judgments of the target brand shift­ ed away from the primed category (contrast effects were observed). The target brand was rated as less (more) expensive when subjects were exposed to extremely expensive (inexpensive) priming stimuli. The magnitude of these priming-induced judgmental shifts was reduced when subjects were highly knowledgeable about the target product domain. In addition to influencing price- and value-related judgments, priming stimuli can facilitate activation of previously encountered product claims (Keller, 1987). For example, a picture of “Little Mikey,” a child actor in a Life cereal television commercial, was placed strategically on packages of Life cereal to facilitate con­ sumer recall of the television commercial while grocery shopping. Moreover, priming can influence what attributes consumers use as a basis for brand com­ parison and evaluation (Bettman & Sujan, 1987). Burke and Srull (1988) examined associative interference in an advertising context. Subjects were exposed to a series of ads for several different brands be­ longing to the same or different product categories. In Experiment 1, an unex­ pected cued recall task was administered for information from ads presented early in the series. In Experiment 2, the cued recall task was administered for infor­ mation from ads presented late in the series. In both studies, cued recall for present­ ed information was greater in varied than in same product context conditions. Hence, cued recall for old information (i.e., information presented in ads that appeared early in the series) was inhibited due to interference from new informa­ tion (i.e., information presented in ads that appeared late in the series). Similar­ ly, cued recall for new information was inhibited due to interference from old information. The former effect was known as retroactive associative interference, and the latter effect was referred to as proactive associative interference. In both cases, the competition between new versus old associations in memory induced retrieval failure (i.e., the inability to locate specific information during memory search). Moreover, retroactive (but not proactive) interference was reduced when subjects were instructed to evaluate brands (vs. ads) prior to ad exposure. In both studies, subjects were unlikely to recall information presented for one brand when asked to recall information presented for another brand. Hence, the observed pat­ tern of results could not be attributed to brand confusion. Experiment 3 examined the effects of ad repetition and associative interfer­ ence on cued recall. In the intense competitive interference condition, subjects were asked to recall information from an ad presented in the middle of a series of ads with ads for directly competing brands presented both before and after the presentation of the target ad (to permit retroactive and proactive interference). In the moderate competitive interference condition, an ad for direct competitor was presented late in the series. In the low competitive interference condition, a varied product context was employed. The target ad was presented once, twice,

420

KARDES

or three times during the series. No effect of target ad repetition on cued recall was observed in the intense competitive interference condition. Cued recall tend­ ed to increase with repetition in the moderate competitive interference condition. Finally, a strong positive relationship between repetition and cued recall was found in the low competitive interference condition. This pattern of results suggests that the shape of the advertising response function depends on the level of competi­ tive advertising that is present. No effect of repetition on recall was observed when competitive interference was intense. Perhaps with greater levels of repeti­ tion, an S-shaped advertising response function may be observed. However, when competitive advertising interference is low, cued recall increases with repetition. Finally, Burke and Srull (1988) suggested that competitive advertising inter­ ference effects should be reduced when consumers are highly knowledgeable about the advertised brands. Recent research has supported this hypothesis (Kent & Allen, 1993). Priming effects also were reduced when context-specific prior knowledge levels were high (Herr, 1989), perhaps because prior knowledge facili­ tated encoding of subsequent information even in difficult learning environments (Alba & Hutchinson, 1987; Johnson & Russo, 1984). The Relationship Between Memory and Judgment

Common sense suggests that there should be a strong relationship between memory and judgment. Consumers should form more favorable product evaluations if they can remember many positive features of a product, as opposed to only a few posi­ tive features. People also should form less favorable product evaluations if they can recall many negative features, as opposed to only a few. However, under many circumstances, different patterns of results are observed for memory ver­ sus judgment. For example, Anderson and Hubert (1963) presented a series of trait adjectives describing a target person and found recency effects in recall (i.e., greater recall for items appearing late in the series) and primacy effects in judg­ ment (i.e., early items were weighed more heavily in impression formation). Con­ sequently, recall and judgment were uncorrelated. Similar results have been observed in many subsequent studies (Hastie & Park, 1986; Lichtenstein & Srull, 1985, 1987). However, on other occasions, memory influences judgment directly and strong recall-judgment correlations are found. For example, memory serves as a basis for judgment when decision makers use the availability or simulation heuristics to generate predictions (Sherman & Corty, 1984; Tversky & Kahneman, 1974). Furthermore, if there is a delay between the time of encoding and the time of judgment, people are likely to refer to memory at the time of judgment. Encod­ ing and judgment are likely to occur at different points in time when processing objectives or goals change over time (Srull & Wyer, 1986). For example, people may judge one aspect of an object or issue during information acquisition, and judge another aspect sometime later (Carlston, 1980; Kardes, 1986). Another

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

421

possibility is that people might gather judgment-relevant information, but keep an open mind and withhold judgment until all of the evidence has been weighed carefully (Kruglanski, 1989). In these cases, information retrieved from memory is likely to be used as a direct input for judgment formation, and, consequently, there is likely to be a strong correlation between recall- and memory-based judg­ ment (Hastie & Park, 1986; Lichtenstein & Srull, 1985, 1987). Information acquisition and judgment formation may occur at different points in time or at the same point in time. People often are exposed to judgment-relevant information sequentially over time. On-line judgment formation occurs when a judgment is formed and updated continuously as judgment-relevant information is acquired (Hastie & Park, 1986; Lichtenstein & Srull, 1985, 1987). Some have referred to this as an anchoring and adjustment process (Hoch & Deighton, 1989; Hogarth & Einhorn, 1992; Tversky & Kahneman, 1974). People form an initial judgment or anchor based on the first pieces of judgment-relevant information that are encountered. This initial judgment is then adjusted or revised as addi­ tional information becomes available. Because adjustment tends to be insufficient, especially when cognitive resources are scarce (Gilbert, Pelham, & Krull, 1988; Trope, 1986), early information is weighed heavily in the final judgment (a prima­ cy effect). Of course, very different processes affect recall for externally presented information (Anderson, 1983), and, consequently, there is likely to be little cor­ respondence between recall and on-line judgment (Hastie & Park, 1986; Lichten­ stein & Srull, 1985, 1987). To test the importance of the distinction between on-line versus memory-based judgment, Lichtenstein and Srull (1985) instructed subjects either to form an evalu­ ation of a product described in an ad (product evaluation conditions) or to rate the ad in terms of how grammatical, coherent, and interesting it was (ad evalua­ tion conditions). Obscure American or European print ads containing large amounts of attribute information were presented. Either 5 minutes or 48 hours following ad exposure, product evaluations and attribute recall were assessed in a second session. Of course, subjects who were instructed to form product evalu­ ations during Session 1 simply could retrieve their previously formed on-line judg­ ments when product evaluations were assessed in Session 2. By contrast, subjects who were told to form ad evaluations during Session 1 did not expect to be asked for product evaluations during Session 2 (i.e., processing goals changed over time). Nevertheless, they could form product evaluations by retrieving attribute information from memory and by forming a judgment based on recalled attri­ butes. As predicted, much greater recall-judgment correlations were obtained in on-line judgment than in memory-based judgment conditions (mean rs = .64 vs. . 22 ).

In a second experiment, subjects again were assigned randomly to on-line or memory-based judgment conditions, and were exposed to unfamiliar ads during Session 1. One hour later, subjects were asked to recall attributes, evaluate products, or perform no task during Session 2. Forty-eight hours after Session 1,

422

KARDES

all subjects were asked to recall attributes during Session 3. Consistent with pri­ or memory research, recall during Session 3 was much greater for subjects who performed a recall task as opposed to no task during Session 2 (the “hypermnesia” effect; Belmore, 1981; Erdelyi & Kleinbard, 1978; Gotz & Jacoby, 1974; Modigliani, 1976). This effect was replicated in on-line and memory-based judg­ ment conditions. However, when subjects performed a judgment task during Ses­ sion 2, recall performance was greater in memory-based (vs. on-line) judgment conditions. This pattern suggests that subjects in memory-based judgment condi­ tions spontaneously recalled attributes in Session 2, resulting in greater recall performance in Session 3 (hypermnesia). That is, memory-based judgment sub­ jects recalled attributes in Session 2 in the absence of explicit instructions to recall attributes in Session 2. Conversely, on-line judgment subjects did not recall at­ tributes spontaneously in Session 2, and, consequently, their recall performance was poor in Session 3. Together, these experiments suggest that memory and judgment are related only when judgments are formed on the basis of informa­ tion retrieved from memory. When judgments are formed on the basis of exter­ nally present information, there is no reason to expect a strong correspondence between memory and judgment (Hastie & Park, 1986). However, Chattopadhyay and Alba (1988) suggested that reliable memoryjudgment correlations should be observed when sufficiently sensitive measures of memory and judgment are provided, regardless of manner of judgment for­ mation. Lichtenstein and Srull (1985, 1987) assessed memory using the “preferen­ tial recall index” (PRI), which is the ratio of the number of positive attributes recalled divided by the number of positive and negative attributes recalled (P/P + N). Chattopadhyay and Alba used the more sensitive summed evaluative im­ plication measure, in which each recalled attribute was weighed by its evaluative and believability ratings (rather than using a simple tally of the number of posi­ tive and negative attributes recalled correctly). Chattopadhyay and Alba assessed judgment using a sensitive 16-item attitude inventory, adapted from Madden, Allen, and Twibble (1988). Chattopadhyay and Alba found a significant memory-judgment correlation when the summed evaluative implication meas­ ure of memory was employed, but not when the less sensitive PRI was employed. Although Chattopadhyay and Alba’s (1988) results appeared to be inconsis­ tent with the two-process model of the relationship between memory and judg­ ment (Hastie & Park, 1986; Lichtenstein & Srull, 1985, 1987), the two-process model should not be rejected prematurely. Chattopadhyay and Alba’s 16-item at­ titude inventory may have induced extremely effortful information processing, resulting in an increase in the subsequent accessibility of the target attribute in­ formation (Tyler et al., 1979) and the target judgment (Kardes, Allen, & Pontes, 1993). Accessible attributes and accessible judgments are likely to be activated simultaneously, resulting in an enhanced memory-judgment relationship. Moreover, the summed evaluative implication measure weighs each recalled at­ tribute with judgments of the attribute, and, therefore, confounds recall with

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

423

judgment. It is well known that judgments of specific attributes are correlated with overall judgments (Anderson, 1967, 1970, 1981, 1982, 1991). Therefore, it is not surprising that the summed evaluative implication measure is correlated with overall judgments. Finally, when different patterns of results are observed for recall versus judgment (e.g., recency in recall and primacy in judgment), weak memory-judgment correlations are observed, even when the sensitive summed evaluative implication measure is employed (Kardes & Herr, 1990). Although explicit memory measures (i.e., recall and recognition) do not al­ ways correlate highly with judgment (Hastie & Park, 1986; Lichtenstein & Srull, 1985, 1987), implicit memory measures may relate frequently to judgment (Nel­ son, Schreiber, & McEvoy, 1992; Roediger, 1990). Although they did not draw the distinction between explicit versus implicit memory, Chattopadhyay and Alba (1988) developed three useful measures of implicit memory: single-fact interpre­ tations, abstractions, and global evaluations. Single-fact interpretations refer to the cognitive representation of single attributes in terms of specific subjective impressions (e.g., the raw data “accelerates from 0 to 60 mph in 7.5 seconds” may be encoded as “fast”). Abstractions refer to the cognitive representation of sets of related attributes in terms of their underlying dimensions (e.g., the raw data “two doors, bucket seats, sleek aerodynamic design” may be encoded as “sporty”). Global evaluations refer to the cognitive representation of raw data in terms of general subjective impressions (e.g., “good car,” “worth considering”). Controlling for recall, reliable abstraction-judgment and global evaluationjudgment relationships were found. Moreover, after a delay, the abstraction-global evaluation model outperformed the traditional counterargument-support argument model of cognitive response. Clearly, information is not always encoded in its raw, and unadulterated form. Single-fact interpretations, abstractions, global evaluations (Chattopadhyay & Alba, 1988), counterarguments, support arguments, source derogations (Petty et al., 1981), and explanations (Hoch, 1984, 1985; Kahneman & Miller, 1986; Sherman, Skov, Hervitz, & Stock, 1981; Sherman, Zehner, Johnson, & Hirt, 1983) often are formed when consumers are exposed to product information. Although the relationship between explicit memory and judgment is fairly well understood (Hastie & Park, 1986; Lichtenstein & Srull, 1985, 1987), the role of implicit memory (Nelson et al., 1992; Roediger, 1990) in judgment continues to provide an interesting and important topic for future research. UTILIZATION OF INFORMATION ABOUT PRODUCTS AND SERVICES When a consumer forms or changes a judgment about a product or a service, the consumer is likely to search the environment and memory for judgment­ relevant information, construe the judgmental implications of this information, and integrate these implications to arrive at a specific judgmental position (Wyer,

424

KARDES

1974; Wyer & Carlston, 1979; Wyer & Srull, 1989). Information implications and processing goals determine the content of a judgment or the position of a judgment along a subjective continuum (Griffin & Tversky, 1992; Kruglanski, 1989; Wyer, 1974; Wyer & Carlston, 1979; Wyer & Srull, 1989). The amount of information used (weight of evidence; Griffin & Tversky, 1992; Koriat, Lichten­ stein, & Fischhoff, 1980), the amount of effort involved in information utiliza­ tion (Kardes, 1993), and motivational factors (i.e., need for closure, need to avoid closure; Kruglanski, 1989) determine the strength of a judgment or the degree of confidence or certainty with which the judgment is held. These general princi­ ples seem to apply to both evaluative (e.g., attitudes, preferences) and nonevaluative (e.g., beliefs, predictions) judgments. Unfortunately, these general principles of judgment have diffused very slow­ ly in the consumer judgment literature, and there are few examples of general theories of consumer judgment. Srull (1982) offered a cogent and compelling explanation for this unfortunate state of affairs. According to Srull, consumer research is primarily phenomenon driven, as opposed to theory driven. Many investigations of consumer judgment focus on a single problem or a single phenomenon, with little emphasis on the development of theory-driven research programs. Furthermore, one large branch of consumer research—the “interpre­ tive” or postmodern branch—explicitly advocates the use of single studies in which a phenomenon is investigated for its own sake, and researcher-improvised theo­ retical frameworks are shunned. However, Lakatos (1974), Lauden (1984), and other philosophers of science have emphasized that little can be learned from a single study, and that true scientific progress depends on the development of in­ tegrated organized multiple-study research programs. Moreover, the phenomenondriven approach implies that all ideas are created equal and that any hypothesis is as good as any other. This results in a disorganized, disjointed, and chaotic field in which almost anything goes. Despite the current state of confusion in the field of consumer behavior, a few useful theories of consumer judgment have emerged. Information Integration Theory

Anderson’s (1967, 1970, 1981, 1982, 1991) information integration theory sug­ gests that attribute ratings provide the building blocks for overall product evalua­ tions. During the valuation stage of the information integration process, consumers evaluate and assign a scale value to each attribute of a target product. During the integration stage, these scale values are combined according to an algebraic integration rule (i.e., the additive rule, the multiplicative rule). Finally, during the output stage, the unobservable integrated impression is translated into an overt response (e.g., a rating on an attitude scale). This cognitive social psychological theory has been used extensively by marketing researchers (Bettman, Capon, & Lutz, 1975; Ford & Smith, 1987; Huber & McCann, 1982; Johnson & Levin,

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

425

1985; Louviere, 1988; Lynch, 1985; Meyer, 1981; Ross & Creyer, 1992; Trout­ man & Shanteau, 1976; Wilkie & Pessemier, 1973) and by marketing practition­ ers (Cattin & Wittink, 1982). Information integration research has focused primarily on the identification of the determinants and types of integration rules people use to combine attribute information. Anderson’s (1970, 1981, 1982) functional measurement theory pro­ vides a powerful framework to achieve this goal. This approach involves presenting product attribute information and asking consumers to provide overall evaluative ratings of the target product. The values of each attribute are manipulated in­ dependently in a factorial design, and an analysis of variance is performed on subjects’ overall ratings. The absence of interactions among the independent vari­ ables suggests that attribute information was combined using an additive integra­ tion rule, covert and overt overall evaluations were related in a linear fashion, and the scales values of each attribute were independent of the scale values of the remaining attributes. Plotting mean overt overall evaluations as a function of the attribute manipulations results in parallel curves. However, the presence of interactions and nonparallel curves is much more difficult to interpret. Non­ parallelism suggests that a nonadditive integration rule was employed, or that a nonlinear relationship existed between covert and overt overall evaluations, or both. The use of experimental procedures that increase the likelihood of a linear relationship between covert and overt overall evaluations reduces the probability of the latter two possibilities. The application of several different monotonic trans­ formation procedures further reduces the likelihood of the latter two interpre­ tations. Information integration theory has been used to analyze several important judg­ mental phenomena, including serial order effects (i.e., primacy and recency ef­ fects in judgment), the negativity effect (i.e., the tendency to weigh unfavorable information heavily in judgment), the positive context effect (i.e., the tendency for overall evaluations to influence ratings of specific attributes), and the set-size effect (i.e., the tendency for judgmental extremity to increase as the amount of information used for judgment increases, even when the implications of each piece of information are controlled). Recent research has suggested that information integration processes also can contribute to the pioneering advantage (i.e., the tendency of consumers to prefer the first brand to enter a market over later entrants). Kardes and Gurumurthy (1992) exposed consumers to attribute information for three brands (A, B, and C) of microwave popcorn. To simulate order of entry in a new market, brand profiles were presented sequentially over a 2-week period. Brand A (or B) was presented in Session 1, and brand B (or A) and brand C were presented 2 weeks later in Session 2. A third session was conducted 2 weeks after the second ses­ sion to examine the persistence of the Session 2 results. The results revealed that consumers learn more attributes of the first brand (vs. followers), even though the amount of information presented for each brand was held constant. Subse­

426

KARDES

quently, judgments of the first brand (vs. followers) were based on a larger amount of information, resulting in a set-size effect. That is, subjects formed more favora­ ble and more confidently held overall evaluations of the first brand (vs. follow­ ers), regardless of whether brand A or B was presented first. Hence, set-size effects can be produced by varying the amount of information presented for each alternative (Anderson, 1981, 1982) or by varying the amount of information learned for each alternative (Kardes & Gurumurthy, 1992). Why do consumers learn more about the pioneering brand? Consider the manner in which consumers are exposed to information about new brands en­ tering a new market. Initially, only one brand is available—the pioneering brand. The features of the pioneering brand are likely to seem novel and interesting to consumers. Novel information is salient, attention-drawing, and likely to be encoded into long-term memory (Anderson, 1983; Kahneman, 1973; Wyer, 1970; Wyer & Carlston, 1979). Rehearsal (Rundus, 1971) and repeated exposure (Madigan, 1969; Nelson, 1977) further increase the amount of information that is likely to be learned about the pioneer. By contrast, the features of follower brands (copy cats) are likely to seem redundant and uninteresting to consumers. Redundant information is nonsalient and is less likely to be encoded into long­ term memory (Anderson, 1983; Kahneman, 1973; Wyer, 1970; Wyer & Carl­ ston, 1979). Consequently, consumers learn more about the pioneer than about followers. If sequential exposure to information about alternative brands is the key de­ terminant of differential learning about the pioneer versus followers, differential learning and the pioneering advantage should be eliminated when information about alternative brands is presented simultaneously. This prediction was con­ firmed in Experiment 2 (Kardes & Gurumurthy, 1992). The pioneering advan­ tage also is known to be weak or nonexistent among late-adopting consumers who fail to consider purchasing a product until the product category is no longer new (Robinson & Fornell, 1985). Of course, late (vs. early) adopters are less likely to be exposed to information about the pioneering brand first. Experiment 2 also included a sequential information presentation condition in which subjects were exposed to information about the first brand incidentally. Specifically, in the first session, subjects received information about the pioneer, but were not asked to form judgments or to recall any information. Instead, after subjects reviewed this information, the experimenter said that he accidentally brought the wrong survey materials with him and he collected the product infor­ mation and dismissed the subjects. Two weeks later, subjects received informa­ tion about two followers and were asked to evaluate all three brands. The results revealed that incidental exposure to information about the pioneer was sufficient to induce learning and preference advantages for the pioneer. Hence, consumers spontaneously learned more about the first brand when attribute information about alternative brands was encountered sequentially.

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

427

The Omission Detection Model

In everyday judgment, people frequently must draw conclusions and make deci­ sions on the basis of limited, unreliable, or incomplete evidence. Rarely is com­ plete information available about all relevant dimensions of a target person, object, or issue. However, sensitivity to omissions (e.g., unmentioned features, partial­ ly described dimensions, or other unknowns) varies systematically as a function of a relatively large set of moderator variables. The omission detection model of judgment based on limited evidence suggests that very different judgmental processes occur when people are insensitive versus sensitive to judgment-relevant omissions (Kardes & Sanbonmatsu, 1993; Sanbonmatsu, Kardes, & Herr, 1992; Sanbonmatsu, Kardes, & Sansone, 1991). When people are insensitive to omissions, missing information is overlooked or neglected and judgments are formed solely on the basis of the available evi­ dence. The omission detection model suggests that neglecting omissions is the default response, because missing information is not very salient or attention draw­ ing and people often are unmotivated or unable to consider all relevant informa­ tion regardless of its availability. When missing information is neglected, people behave as if the available set of information represents a complete set of infor­ mation. When an incomplete set of information is treated as if it were complete, inappropriately extreme and confidently held judgments are formed (Kardes & Sanbonmatsu, 1993; Sanbonmatsu et al., 1991, 1992). By contrast, when people are sensitive to omissions, people are more likely to be cognizant of the fact that their judgments are based on limited or incom­ plete evidence. This insight produces uncertainty and induces judgmental cau­ tion. One way to reduce this uncertainty is to generate elaborative inferences that enable consumers to fill in missing details by going beyond the information given (Kardes, 1993). Generally, these inferences are moderate, because people are often reluctant to draw strong conclusions in the face of a lack of direct evi­ dence (Kardes, Kim, & Lim, in press; Sanbonmatsu et al., 1991). In other cases, inferences are not drawn, because missing information is expected to be uncorrelated with presented information (Kardes & Sanbonmatsu, 1993) or be­ cause an appropriate inferential rule is inaccessible from memory (Dick, Chakravarti, & Biehal, 1990; Holland, Holyoak, Nisbett, & Thagard, 1986). When inference generation is unlikely, judgments are adjusted to allow for un­ certainty. This adjustment usually involves a shift toward a relatively neutral po­ sition, because a moderate stance is easy to justify and is updated readily as additional information becomes available (Cialdini, Levy, Herman, & Evenbeck, 1973; Jaccard & Wood, 1988; Sanbonmatsu et al., 1991, 1992; Simmons & Lynch, 1991). Moreover, when allowances are made for informational uncer­ tainty, overconfidence is reduced (Griffin, Dunning, & Ross, 1990; Sanbonmat­ su et al., 1991, 1992).

428

KARDES

One way to vary the relative completeness versus incompleteness of informa­ tion is by manipulating set size or the amount of information available for judg­ ment. Prior research has shown that between-subjects set-size effects rarely are found, whereas within-subjects set-size effects frequently are observed (Ander­ son, 1981; Sloan & Ostrom, 1974). However, the omission detection model sug­ gests that between-subjects set-size effects should be observed reliably when consumers are sensitive to omissions, and that within-subjects set-size effects should be reduced when consumers are insensitive to omissions. In betweensubjects designs, respondents are exposed to one and only one set of informa­ tion. When omissions are neglected, even a small set of information is likely to be treated as if it were complete. Consequently, extreme and confidently held judgments are formed in both small and large set-size conditions. However, respondents possessing a high level of context-specific prior knowledge are like­ ly to be sensitive to omissions, and, consequently, high-knowledge respondents should form more moderate judgments in small set-size conditions. Consistent with this prediction, reliable between-subjects set-size effects are observed for high-knowledge respondents, but not for moderate or low-knowledge respondents (Sanbonmatsu et al., 1992). In within-subjects designs, respondents are exposed to two objects—one described by a large set of information (e.g., eight attributes) and one described by a small set of information (e.g., six attributes). Respondents should be rela­ tively sensitive to omissions when they can readily see that one brand is described by a large set of information (the large set brand) and the other is described by a small set (the small set brand). Consequently, within-subjects set-size effects commonly are observed (Anderson, 1981; Sloan & Ostrom, 1974). However, when people compare two objects, one object (the subject of comparison) serves as the starting point and sets the agenda for the comparison process and the other object (the referent of comparison) is less focal (Sanbonmatsu, Kardes, & Gib­ son, 1991; Tversky, 1977). When the large (vs. small) set brand is more focal, the presence of additional features in the large set brand and the absence of these features in the small set brand is highlighted. Conversely, when the small set brand is more focal, features present for the large and absent for the small set brand are less prominent. Consequently, within-subjects set-size effects are reduced when the large set brand serves as the referent (vs. subject) of compari­ son, because sensitivity to omissions is reduced (Kardes & Sanbonmatsu, 1993). The finding that consumers are typically insensitive to omissions is consistent with a broader set of findings in the psychology literature. It is well known that nonoccurrences and nonevents are neglected in judgments of contingency or covar­ iation (Klayman & Ha, 1987; Nisbett & Ross, 1980). Furthermore, features deleted from an object over time are surprisingly difficult to specify, even when people recognize that something is different about the object (Agostinelli, Sherman, Fa­ zio, & Hearst, 1986). Moreover, animals and people learn an association between two events less readily when the signal is the absence (vs. presence) of a cue

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

429

(the feature-positive effect; Newman, Wolff, & Hearst, 1980). Finally, decision makers neglect items not presented in a fault tree, regardless of whether a full or a pruned (less complete) fault tree is provided (Fischhoff, Slovic, & Lichten­ stein, 1978; Hirt & Castellan, 1988). However, several variables moderate sensitivity to omissions, including missing information cues, prior knowledge, initial processing goals, and direction of com­ parison. As sensitivity to omissions increases, judgmental extremity and over­ confidence decreases. This pattern has been observed consistently across a wide range of judgments, including attitudinal judgments (Sanbonmatsu et al., 1992), preference judgments (Kardes & Sanbonmatsu, 1993), inferential judgments (San­ bonmatsu et al., 1991), and likelihood judgments (Peterson & Pitz, 1988). Fu­ ture research should attempt to identify additional variables that moderate sensitivity to omissions (e.g., need for closure, need to avoid closure, accounta­ bility, time pressure) and should investigate the extent to which sensitivity to omis­ sions mediates judgmental extremity and confidence in other types of judgments (e.g., judgments of category membership, causal inferences, trait inferences). A clearer understanding of the psychological processes involved in omission de­ tection should facilitate the development of decision aids for improving judgment based on limited evidence. Data-Driven Versus Theory-Driven Inference Processes

Attribute-based, stimulus-based, or data-driven judgment processes are likely when information is analyzed into discrete units and when the implications of these dis­ crete units are considered separately (Wyer & Carlston, 1979). The classic ex­ ample of data-driven judgment processes was provided by Anderson’s (1981, 1982) information integration theory. Consumers are especially likely to engage in datadriven processes when (a) no prior judgment of a target product exists in long­ term memory (e.g., the product is pioneering or novel), (b) information is of different types or is difficult to integrate (e.g., a mixture of visual and verbal information in an ad), (c) information is provided by different sources (e.g., a sales representative and a Consumer Reports article), or (d) information pertains to different objects (e.g., two or more alternative brands; Wyer & Carlston, 1979). Because data-driven analyses require a relatively high level of cognitive ef­ fort, several models of judgment suggest that consumers attempt to apply prior knowledge and experience to a judgment task before they engage in more effort­ ful data-driven information processing (Brewer, 1988; Fiske & Neuberg, 1990; Fiske & Pavelchak, 1986; Wyer & Carlston, 1979; Wyer & Srull, 1989). Con­ sumers’ prior knowledge and experiences can be conceptualized as idiosyncratic “theories” specific to individual consumers. Under some circumstances, consumers focus primarily on externally present data and engage in effortful data-driven processes. On other occasions, consumers focus primarily on previously formed theories stored in long-term memory and engage in less effortful memory-based,

430

KARDES

theory-driven processes. These theories can be as simple as a single attitudinal judgment stored in long-term memory (e.g., if I know I like Pepsi, I do not need to engage in effortful data-driven analyses each time I wish to purchase a soft drink), or as complex as an elaborate organized prior knowledge structure or sche­ ma. Chessmasters, experienced consumers, and other experts are so knowledge­ able that they often do not need to pay much attention to externally present details; these details already are stored in long-term memory (Alba & Hutchinson, 1987; Hastie, 1981; Johnson & Russo, 1984; Wyer & Carlston, 1979; Wyer & Srull, 1989). Moreover, retrieving previously formed judgments and schemata from memory is more efficient than recomputing judgments and schemata each time this information is needed to perform a cognitive task. Alba and Hutchinson (1987) distinguished between four types of theory-driven inference processes: evaluation-based (Beckwith & Lehmann, 1975; Sawyer, 1988; Sawyer & Howard, 1991), correlation-based (Dick et al., 1990; Ford & Smith, 1987; Huber & McCann, 1982; Johnson & Levin, 1985; Kardes & San­ bonmatsu, 1993; Meyer, 1981; Ross & Creyer, 1992), category-based (Fiske & Pavelchak, 1986; Hong & Wyer, 1989, 1990; Sujan, 1985; Sujan & Dekleva, 1987), and schema-based inference processes (Hastie, 1983; Kardes, 1988a, 1988b, 1993; Kardes, Kim, & Lim, in press; Stayman & Kardes, 1992). The simplest type of inference, the evaluation-based inference, involves using a previ­ ously formed summary judgment (e.g., an overall product evaluation) retrieved from memory to serve as a basis for a specific judgment (e.g., an attribute rat­ ing) or for a related summary judgment. For example, a judgment bearing on the suitability of an individual for one occupation (e.g., physician) can be used to infer the suitability of the individual for another related occupation (e.g., law­ yer) without requiring reanalysis of specific trait information (Lingle & Ostrom, 1981). Similarly, an overall impression of the relative quality of an advertised product can be used to infer the suitability of the product for purchase without requiring reanalysis of specific product attribute information (Sawyer, 1988; Saw­ yer & Howard, 1991). If two product attributes are believed to be correlated, and if information is available for one attribute but not the other, consumers can infer a value for the missing attribute on the basis of the value of the available attribute. For example, many consumers believe that there is a strong relationship between price and qual­ ity (Huber & McCann, 1982; Johnson & Levin, 1985; Meyer, 1981). Conse­ quently, if consumers know that a target product is very expensive, but are unsure about the quality of the product (e.g., perhaps quality is ambiguous or difficult to judge), many consumers will infer that the product is of high quality. Con­ versely, if consumers know that a target product is very inexpensive, but quality is unknown, many consumers will assume that the product is of low quality. Of course, these inferences are contingent on the belief that price and quality are correlated, and this belief may vary as a function of individual differences in context-specific prior knowledge.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

431

Category-based inference processes simplify judgment and decision making by enabling people to refer to prior knowledge about a list of attributes (as op­ posed to a pair of attributes) common to exemplars or instances of a given group or category whenever a member of the target category is encountered. Category activation and utilization is especially likely when (a) the category has been primed (Herr, 1989; Higgins, Rholes, & Jones, 1977; Srull & Wyer, 1979, 1980), (b) a typical (vs. atypical) member of the category is encountered (Fiske & Neuberg, 1990; Fiske & Pavelchak, 1986; Sujan, 1985; Sujan & Dekleva, 1987), (c) involvement is low (Brewer, 1988; Maheswaran et al., 1992), or (d) when a very complex judgment or decision is required (Bodenhausen & Lichtenstein, 1987; Bodenhausen & Wyer, 1985). Examples of category-based inference processes using subordinate (i.e., a specific brand name, such as Nabisco or Sony), basic (i.e., a product class, such as crackers or audio equipment), and super­ ordinate level categories (i.e., stereotypes, such as Japanese people or Japanesemade products) are provided next. Consumers are especially sensitive and responsive to brand names when in­ volvement is low, and, consequently, subordinate-level category-based inference processes are likely (Maheswaran et al., 1992). When involvement is high, con­ sumers engage primarily in data-driven judgment processes, especially when the implications of brand names and important attributes are evaluatively consistent. However, when involvement is high and when these implications are inconsis­ tent (i.e., when the judgment task is relatively complex), both sets of implica­ tions are integrated into overall product evaluations. Hence, the extent to which consumers rely on brand names during product evaluation depends on initial processing goals and on the complexity of the judgment task. This study has im­ portant implications for understanding the conditions under which brand exten­ sions are likely to be successful. Brand extensions enable firms to reduce the risk of marketing new products by attaching established brand names to the new product offerings (Aaker & Keller, 1990; Farquhar, 1989; Farquhar, Herr, & Fazio, 1990; Kardes & Allen, 1991; Park, Milberg, & Lawson, 1991). This strategy is likely to be effective when involvement is low, judgment complexity is high (Maheswa­ ran et al., 1992), the new product is similar to other products sharing the parent brand name (Fiske & Pavelchak, 1986), and when a brand name is likely to prime a product category and vice versa (Farquhar et al., 1990). Of course, brands differ in their effectiveness as priming stimuli or cues for category activation. General­ ly, effectiveness increases as cue set size (i.e., the number of different cognitive associations linked to a brand name) decreases or as target set size (i.e., the number of different brands linked to a product category) decreases (Meyers-Levy & Tybout, 1989; Nelson et al., 1992). Theory-driven processes can occur for general product categories (i.e., product classes, such as soft drinks) as well as for specific brand categories (e.g., Coca Cola). Sujan (1985) exposed subjects to product descriptions in which product labels (35 mm or 110 camera) and product attributes (35 mm or 110 camera

432

KARDES

attributes) were manipulated independently. In half of the conditions, the attri­ butes matched the label (e.g., a 35 mm camera was described as having in­ terchangeable lenses, great versatility, and many accessories), whereas in the remaining conditions, a mismatch occurred (e.g., a 35 mm camera was described as small, pocketable, and easy to use). When the label and attributes matched, it was predicted that subjects would perceive the product as typical for the product category, and that subjects would engage in category-based judgment processes. Conversely, attribute-based judgment processes were expected in mismatch con­ ditions. Consistent with Fiske and Pavelchak’s (1986) dual-process model, category-based (as opposed to attribute-based) judgment processes resulted in faster evaluation times, more category-related verbal protocols, and fewer attributerelated verbalizations. Moreover, these effects were more pronounced for sub­ jects who were knowledgeable (vs. unknowledgeable) about the target product category (i.e., cameras). Recent research on country-of-origin effects on product evaluations has sug­ gested that country-of-origin information (e.g., “Made in Japan”) activates a stereo­ type and produces effects conceptually similar to those produced by stereotypes in social perception. Hong and Wyer (1989) presented country-of-origin and at­ tribute information about two products (a personal computer and a video cassette recorder) and asked subjects either to read the information and consider its clari­ ty (comprehension set) or to form an impression of the described product (im­ pression set). The country-of-origin information had either favorable (i.e., Japan, West Germany) or unfavorable (i.e., Mexico, South Korea) evaluative implica­ tions, and this information was presented either before or after the attribute in­ formation. Attribute valence also was manipulated. No assimilation effects and no serial order effects were observed. If purely theory-driven processes were oper­ ative, product evaluations should have been influenced strongly by country of origin, and attribute information should have been ignored or neglected. However, the results implied that country-of-origin information influenced product evalua­ tions directly by providing an input for evaluation, and that country-of-origin in­ formation also affected product evaluations indirectly by stimulating greater levels of cognitive elaboration of attribute information (provided that an impression set had not stimulated a high level of cognitive elaboration already). In a follow-up study, Hong and Wyer (1990) manipulated the time interval between the presentation of country-of-origin versus attribute-information (either a 24-hour delay or no delay was used). Again, order of presentation, attribute valence, and the evaluative implications of the country-of-origin information were manipulated orthogonally. When country-of-origin information was presented 24 hours prior to the presentation of attribute information, country of origin had a direct impact on product evaluations and an indirect effect due to the assimila­ tion of the implications of attribute information toward the implications of the country-of-origin information. By contrast, in no delay conditions, purely attributebased processing occurred and country-of-origin information was treated no

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

433

differently than any other piece of attribute information. Together, the Hong and Wyer (1989, 1990) studies suggested that stereotype activation results in direct and indirect effects of country-of-origin information on product evaluations. However, the extent to which these processes generalize to social perception is unclear. To summarize, category-based inference processes enable consumers to uti­ lize information stored in memory and prevents consumers from having to de­ liberate extensively over externally presented details. This is generally the case regardless of whether low- (e.g., brand names; Maheswaran et al., 1992), medium- (e.g., product classes; Sujan, 1985), or high-level (e.g., stereotypes; Hong & Wyer, 1989, 1990) categories are activated from memory. However, several studies have shown that judgmental implications derived from activated categories are not used exclusively, but are used in conjunction with the judg­ mental implications of attribute information (Hong & Wyer, 1989, 1990; Ma­ heswaran et al., 1992; Sujan, 1985). Nevertheless, purely category-driven inference processes may occur when the costs of judgmental inaccuracy are very low, when attribute information is very technical (perhaps beyond a consumer’s comprehension), when several alternative brands are evaluated (i.e., when in­ formation load is very high), or when the costs of indecisiveness are high (Kruglan­ ski, 1989). Wyer and Gordon (1984) suggested that categories are represented in memory as unorganized lists of attributes, whereas schemata possess an internal structure consisting of formal links and associations between attributes. Hastie (1983) dis­ tinguished between context-dependent schemata, which are formed on the basis of the idiosyncratic experiences of individual consumers (e.g., someone who lives in New York City is likely to have a more complex cognitive map of the city, compared with a visitor); and context-independent schemata, which are abstract rules of human reasoning that can be applied in virtually any setting. Although a number of abstract inferential rules have been identified (Holland et al., 1986; Smith, Langston, & Nisbett, 1992), I limit my discussion to one illustrative ex­ ample, the propositional logic rule modus ponens. The basic structure of this rule is “if p then q; p therefore q.” Several theorists suggest that if-then rules provide the building blocks of context-independent schemata (Anderson, 1983; Hastie, 1983; Holland et al., 1986; Kruglanski, 1989; Wyer & Carlston, 1979). Stayman and Kardes (1992) examined schema-driven inference processes in an advertising context. Subjects were exposed to a mock ad containing four sets of logical arguments using if-then structures to relate features to benefits. For example, one set stated “The CD880 has an audible fast forward and reverse. Finding a specific passage is easier with an audible scanner. The CD880 makes it easy to find a specific passage.” The last sentence was presented explicitly in the ad in explicit conclusion conditions, whereas it was omitted from the ad in implicit conclusion conditions. It should be emphasized that the omitted conclu­ sion can be inferred by all subjects, regardless of their level of prior knowledge

434

KARDES

concerning compact disc players. The application of abstract context-independent inferential rules does not require context-specific prior knowledge. Of course, logical reasoning cannot be performed effortlessly and subjects must think carefully about the arguments to infer the appropriate conclusions. However, not all subjects are motivated sufficiently to engage in inferential reasoning. Study 1 indicated that subjects high (vs. low) in need for cognition (i.e., subjects who enjoy performing effortful cognitive tasks; Cacioppo & Petty, 1982) were more likely to generate inferences spontaneously. Likelihood of spontaneous inference generation was assessed through the use of a response latency procedure that in­ volves the comparison of subjects’ latencies to inferential inquiries across three conditions: conditions in which spontaneous inference generation was expected, conditions in which spontaneous inference generation was unexpected, and con­ ditions in which the appropriate conclusions were presented explicitly and there­ fore could not be inferred (see Kardes, 1988b). Study 1 also demonstrated that if inferences were generated spontaneously, low (vs. high) self-monitoring subjects (i.e., subjects sensitive to self-generated beliefs; Snyder, 1974) were more likely to use these inferences as inputs for product evaluation. Likelihood of spontaneous inference utilization was assessed by measuring evaluation latencies. Evaluations formed on the basis of effortful inferential reasoning processes were more accessible from memory and were ex­ pressed more quickly than evaluations formed on the basis of less effortful cog­ nitive processes (Kardes, 1988b, 1993). Study 2 showed that spontaneous inference generation was more likely when involvement was high (vs. low). Again, spon­ taneous inference utilization (given generation) was more likely for low (vs. high) self-monitoring individuals. Together, these studies suggested that creating con­ ditions that encourage spontaneous inference generation and that facilitate spon­ taneous inference utilization is critical for fostering the development of strong brand attitudes. A clearer understanding of the complex processes involved in abstract context-independent inferential reasoning is necessary to achieve this goal.

The Availability-Valence Model

Kisielius and Stemthal’s (1984, 1986) availability-valence model of consumer judg­ ment suggested that product evaluations are determined by the readiness or ease with which judgment-relevant information can be retrieved from memory, and by the valence or favorableness of retrieved information. The term “availability” was borrowed from Tversky and Kahneman’s (1974) “availability heuristic,” which is a simple rule of thumb for judging frequency or likelihood on the basis of the ease with which instances or examples of a category come to mind. However, the term availability is actually a misnomer. Tulving and Pearlstone’s (1966) useful distinction between availability (i.e., whether an item is stored in long­ term memory) and accessibility (i.e., the readiness or ease with which available

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

435

information is retrieved from long-term memory) was accepted widely in social cognition (Fiske & Taylor, 1991; Wyer & Srull, 1989). Hence, the availabilityvalence model is referred to as the accessibility-valence model throughout the remainder of this chapter to maintain consistency with the social cognition litera­ ture and to maintain consistency within the chapter. In one of the first investigations of the accessibility-valence model, subjects received a set of prosecution arguments and a set of defense arguments bearing on a hypothetical drunk driving case (Reyes, Thompson, & Bower, 1980). Although the implications of these arguments were held constant, the manner in which these arguments were presented was manipulated experimentally. Half of the subjects received vivid (i.e., concrete and image-provoking) prosecution ar­ guments and pallid (i.e., abstract and boring) defense arguments, whereas the remaining subjects received pallid prosecution arguments and vivid defense ar­ guments. For instance, one vivid prosecution argument stated that the defendant knocked over a bowl of guacamole dip and splattered guacamole on a white shag carpet (implying that he or she was drunk). The pallid version of the same argu­ ment stated simply that he or she knocked a bowl to the floor. One vivid defense argument implied that the garbage truck the defendant hit was gray, dull, and hard to see. The garbage truck driver asked “What do you want, I should paint it pink?” The pallid version of the same argument omitted any reference to a pink garbage truck. Subjects’judgments of the guilt of the defendant were uninfluenced by these manipulations immediately after the evidence was presented. However, after a 48-hour delay, subjects in vivid prosecution arguments conditions recalled more prosecution arguments and judged the defendant as more likely to be guilty. Subjects in vivid defense argument conditions recalled more defense arguments and judged the defendant as more likely to be innocent. Thus, judgments of guilt versus innocence depend on the relative accessibility from memory of prosecutionversus defense-related arguments. Kisielius and Sternthal (1984) extended these findings to consumer judgment by presenting favorable product claims pertaining to a new brand of shampoo. When subjects simply read these claims relatively favorable product evaluations were formed. However, when subjects elaborated on these claims, either because they were instructed to elaborate or because vivid elaboration-inducing pictures were presented along with the claims, less favorable product evaluations were formed. Because subjects’ self-generated elaborations were less favorable and more accessible, relative to the implications of the product claims taken at face value, less favorable overall evaluations were formed when elaboration was likely (vs. unlikely). The results of a second experiment indicated that the effects of the elaboration-inducing pictures were reduced when the information was presented quickly (vs. slowly), because a fast rate of presentation reduces the opportunity to elaborate. A third experiment demonstrated that elaborating on pictorial infor­ mation uses up cognitive capacity, leaving insufficient resources for processing additional extracommunication information. Consequently, this additional infor­

436

KARDES

mation had a weaker effect on product evaluations when pictorial information was presented (vs. omitted). Hence, many different manipulations influence in­ formation accessibility, and the effect of any given piece of information on judg­ ment increases as its accessibility from memory increases. The dual-process model of the relationship between memory and judgment (Hastie & Park, 1986; Lichtenstein & Srull, 1985, 1987) suggests an important qualification to the role of accessibility in judgment: accessibility should have a strong impact on memory-based judgments, but no impact on on-line judgments. However, contrary to this prediction, Shedler and Manis (1986) obtained relia­ ble vividness effects on judgment even when on-line judgmental processes were likely. Moreover, strong vividness-memory and vividness-judgment correlations were found, but no direct relationship between memory and judgment was ob­ served. Shedler and Manis explained their findings by suggesting that vividness manipulations produce independent effects on memory and judgment, and these effects result in a spurious correlation between memory and judgment when one fails to control for indirect effects using path-analytic procedures. However, con­ sistent with the accessibility-valence model, recent research has indicated that a direct effect of memory on judgment is found with implicit memory (i.e., memory for self-generated elaborations; Roediger, 1990), but not with explicit memory (i.e., recognition, recall) measures (Moser, 1992). A useful model of consumer judgment should apply to many different judg­ mental phenomena, rather than to a single phenomenon (e.g., the vividness ef­ fect). Sternthal and his colleagues showed that the accessibility-valence model provides a useful integrative framework for explaining a surprisingly diverse ar­ ray of judgmental phenomena. For example, when a valid message is presented by a biased or otherwise untrustworthy communicator, little persuasion occurs initially. However, persuasion tends to increase with the passage of time (the sleep­ er effect). The availability-valence model suggests that, under some conditions, message accessibility from memory remains high and communicator accessibility decreases over time (Hannah & Sternthal, 1984; Mazursky & Schul, 1988; Schul & Mazursky, 1990). Consequently, the effect of the message on persuasion re­ mains high, whereas the effect of the communicator on persuasion decreases over time. The accessibility-valence model also suggests that the effects of extrinsic in­ centives, such as coupons for purchasing a product at a reduced price, either can increase or decrease the favorableness of product evaluations (Tybout & Scott, 1983). Prior research has shown that direct behavioral experience with a product (e.g., tasting a new beverage) results in strong, well-defined brand attitudes that are highly accessible from memory, whereas indirect experience (e.g., second­ hand information) results in weak, poorly defined attitudes that are less accessi­ ble from memory (Fazio & Zanna, 1981). Favorable attitudes that are highly accessible from memory become even more favorable when coupons are provid­ ed (vs. not provided), because extrinsic incentives supply additional reasons for

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

437

purchasing a target product. By contrast, favorable attitudes that are inaccessible from memory become less favorable when coupons are provided (vs. not provid­ ed), because extrinsic incentives undermine poorly defined attitudes. The latter phenomenon often is referred to as the overjustification effect, because addition­ al reasons for engaging in a target behavior reduce the accessibility of the origi­ nal reason, the initial attitude (Bern, 1972; Tybout & Scott, 1983). However, the overjustification effect is eliminated when a reminder is presented to increase the accessibility of the initial attitude (Fazio, 1981). Effects of multiple request techniques on consumer compliance also can be understood in terms of the implications of the accessibility-valence model (Tybout, Sternthal, & Calder, 1983). People are more likely to comply with a moderately expensive or effortful target request when it is preceded by an inexpensive or trivial initial request (the foot-in-the-door effect; Cann, Sherman, & Elkes, 1975; Cialdini, 1988; Freedman & Fraser, 1966). They are also more likely to comply with such a request when it is preceded by an unreasonably expensive or effort­ ful initial request (the door-in-the-face effect; Cann et al., 1975; Cialdini, 1988; Cialdini et al., 1975). Hence, people are more likely to comply with a target re­ quest when it is preceded by a small (foot in the door) or large (door in the face) initial request (as opposed to no initial request). The accessibility-valence model suggests that the accessibility of the subject’s response to an initial request, the accessibility of the requester’s response, and the favorableness of these responses with respect to subsequent compliance with the target request can vary. The subject’s response to an initial request increases in accessibility as the amount of effort involved in responding to the initial re­ quest increases, or as the delay between multiple requests decreases. Compli­ ance with the initial request has favorable implications for subsequent compliance, whereas noncompliance with the initial request has unfavorable implications for subsequent compliance. Hence, ideally, the initial request should be large enough to ensure that the subject’s response will be accessible, but small enough to en­ sure that the subject’s response will be favorable. The requester’s response is more accessible when multiple requests are delivered by the same, as opposed to a differ­ ent, individual, or when the target request immediately follows the initial request. Deescalating requests (i.e., a large request followed by a small request) have favorable implications for subsequent compliance, because the requester appears less demanding and more reasonable. By contrast, escalating requests (i.e., a small request followed by a large request) have unfavorable implications for subsequent compliance, because the requester appears more demanding and less reasonable. Consistent with the accessibility-valence interpretation of multiple-request effects, the foot-in-the-door effect increases in magnitude as the amount of effort required to comply with the initial request increases, provided that compliance with the initial request is obtained. The foot-in-the-door effect also is more pronounced when multiple requests are delivered by the same requester and when the target request is delivered following a delay. Conversely, the door-in-the-face effect

438

KARDES

is greater when multiple requests are delivered consecutively by the same individu­ al (Tybout et al., 1983). Increasing the accessibility of favorable information enhances the judgmental impact of this information. Accessibility increases with frequent, recent, or elaborative processing. By contrast, decreasing the accessibility of unfavorable information should reduce its judgmental impact. This can be done by inducing proactive or retroactive associative interference. Tybout, Calder, and Sternthal (1981) tested the effectiveness of an interference-inducing procedure for com­ bating the rumor that McDonald’s used worms in their hamburgers. Although most consumers did not believe the rumor, an association in memory between McDonald’s and worms brings worms to mind every time a McDonald’s outlet is encountered. Simply thinking about worms can cause a decrease in appetite, even if the rumor is known to be false. McDonald’s actually denied the rumor to no avail, and this effect was replicated in the laboratory of Tybout et al. (1983). However, in the associative interference condition of this study, subjects com­ pleted a lengthy survey containing several questions about many different aspects of McDonald’s. Responding to these questions activated many associations regard­ ing McDonald’s other than the worm association. Consequently, the probability of retrieving the worm association when subjects thought about McDonald’s de­ creased, and the impact of the rumor on evaluations of McDonald’s also decreased. One final point about the accessibility-valence model should be emphasized. Recent research has shown that elaborative processing can increase or decrease recall performance, depending on whether reconstructive inference processes are facilitated or inhibited, respectively (Walker, 1986). This finding was inconsis­ tent with Anderson’s (1983) hypothesis that elaborative processing produces mul­ tiple pathways for directly accessing a specific piece of information stored in long-term memory. Instead, elaborative processing appears to provide multiple cues for generating items that cannot be retrieved directly. Multiple cues enhance recall performance when reconstructive inference processes are likely. In con­ trast, multiple cues reduce recall performance (due to associative interference) when reconstructive inference processes are unlikely. Hence, the effects of elabora­ tive processing on memory and judgment are more complex than the accessibilityvalence model implies. Nevertheless, the accessibility-valence model provides a powerful integrative framework for explaining a wide variety of interesting and important judgmental phenomena. The Accessibility-Diagnosticity Model

Feldman and Lynch’s (1988) accessibility-diagnosticity model suggests that the likelihood with which a particular piece of information stored in long-term memory is used as an input for memory-based judgment or choice depends on (a) the ac­ cessibility of the information (i.e., the ease with which the information can be retrieved from memory), and (b) the diagnosticity or relevance of the informa­

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

439

tion. Like the availability-valence model, the accessibility-diagnosticity model suggests that any factor that increases the relative accessibility of a given piece of information also increases the likelihood that the information will be used as an input for memory-based judgment. Correspondingly, any factor that decreases the relative accessibility of a particular piece of information also decreases the likelihood that the information will be used in judgment. Moreover, the direction (i.e., positive vs. negative) and the extremity (i.e., very positive/negative vs. somewhat positive/negative) of a judgment depends on the evaluative implica­ tions (i.e., valence) of accessible evidence. The accessibility-diagnosticity model extends the availability-valence model by emphasizing the key role of information diagnosticity or relevance in judg­ ment. Accessible information is unlikely to be used as an input for judgment if this information is perceived as nondiagnostic or irrelevant with respect to the attainment of a salient processing objective. Information diagnosticity is assessed in an on-line manner as the information is retrieved from memory. If a single piece of information is perceived to be sufficient for achieving a particular judg­ mental objective, memory search ceases and a judgment is formed solely on the basis of this information (Wyer & Srull, 1986, 1989). If currently accessible in­ formation is perceived as nondiagnostic or insufficient for achieving a process­ ing goal, memory search continues until enough information for achieving the goal has been accumulated. Of course, the amount of information that is consid­ ered sufficient for a judgment varies as a function of the scope of the processing goal (e.g., choosing the best brand requires more information than choosing a satisfactory brand). Highly diagnostic evidence implies one judgmental position or one choice al­ ternative to the exclusion of all others. For instance, if a consumer’s goal is to buy the least expensive brand from a small set of brands retrieved from memory, only one piece of information (i.e., price) is sufficient to achieve this objective, and all other information is nondiagnostic or irrelevant. Other goals may require more extensive information. Nondiagnostic evidence implies many possible judg­ mental positions or many different choice alternatives. For example, price infor­ mation is nondiagnostic if all brands in the consideration set are equivalent in price. In this case, some criterion other than price must be used to select one brand from the set of alternatives. Similarly, information bearing on the number of wheels of a new automobile (i.e., four) implies that the model may belong to virtually any automobile category (e.g., luxury car, sports car, sedan, station wagon), and, hence, this information is nondiagnostic with respect to the goal of categorizing the new model. Information that implies multiple possible con­ clusions, categorizations, interpretations, hypotheses, or judgmental positions is not very useful for achieving a particular judgmental objective. In one of the first experimental investigations of the accessibility-diagnosticity model, Lynch, Marmorstein, and Weigold (1988) exposed subjects to attribute information bearing on three brands of color televisions: Sanyo, Philco, and Sharp

440

KARDES

(Session 1). After evaluating these three brands, subjects received attribute in­ formation about four distractor brands (Session 2). In high-interference condi­ tions, subjects were instructed to memorize distractor brand information, whereas in low-interference conditions, subjects performed an unrelated task. Next, sub­ jects were asked either to choose between a target memory brand (Sanyo) and a stimulus brand (Sylvania), to recall Sanyo attributes, to recall their prior evalu­ ations of Sanyo, or to perform an unrelated task. Two days later, subjects were asked to recall Sanyo’s attributes and also their earlier evaluation of Sanyo (Ses­ sion 3). The results indicated that when attribute information was accessible (i.e., lowinterference conditions), subjects used recalled attributes as inputs to choice (i.e., subjects engaged in attribute-based choice), and failed to use recalled evaluations as inputs to choice. Attribute recall during the third and final session was as high for subjects who were asked to perform the choice task during the second session as for subjects who were asked to recall attributes during the second session. This pattern suggested that choice task subjects spontaneously recalled attributes from memory even though they were not asked explicitly to do so (Lichtenstein & Srull, 1985). Consequently, choice task and attribute recall task (vs. unrelated task) subjects exhibited superior attribute recall performance in the final session (the hypermnesia effect; Belmore, 1981; Erdelyi & Kleinbard, 1978; Gotz & Jacoby, 1974; Modigliani, 1976). Different results occurred in high-interference condi­ tions, however. In these conditions, attribute recall in the third session was less when subjects had been asked to perform the choice task during the second ses­ sion than when they had been asked to recall attributes. This suggests that choice task subjects made no attempt to recall attribute information when this informa­ tion was relatively inaccessible from memory. In a follow-up experiment, Lynch et al. (1988) examined the effects of the perceived diagnosticity of prior evaluations on the use of these evaluations ver­ sus recalled attributes as a basis for choice. Subjects received attribute informa­ tion pertaining to several brands from several different product categories. Then they were asked to rate each brand in terms of both “quality without considering price” and “value for the money.” The perceived diagnosticity of prior evalua­ tions was operationalized in terms of the degree of consistency between quality and value ratings. The perceived diagnosticity of subjects’ prior evaluations was assumed to be high when both of these ratings implied that the same brand should be chosen and to be low when the ratings were inconsistent (i.e., when one brand was rated high in quality, but low in value, and another brand was rated low in quality, but high in value). As predicted, prior evaluations were used as inputs to choice when their perceived diagnosticity was high, whereas recalled attri­ butes were used as inputs to choice when the perceived diagnosticity of prior evaluations was low. The accessibility-diagnosticity model provides a useful framework for under­ standing the vividness effect (Herr et al., 1991). Vividly presented information

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

441

captures attention and encourages elaborative processing. Consequently, vivid information is retrieved readily from long-term memory. Although a particular piece of information is more likely to be used as an input for judgment or choice as its accessibility from memory increases, the accessibility-diagnosticity model emphasizes that accessibility does not guarantee utilization. Nondiagnostic in­ puts are not likely to be used, even if they are highly accessible. Moreover, ac­ cessible and highly diagnostic inputs are likely to reduce the judgmental effects of equally accessible, but less diagnostic, inputs. Consistent with these predic­ tions, Herr et al. (1991) found that vividly presented information is more acces­ sible and is weighted more heavily in judgment, relative to equally probative, but pallidly presented, information. However, the vividness effect is reduced or eliminated when other more diagnostic inputs are available for judgment. Spe­ cifically, prior beliefs tend to be perceived as highly diagnostic and people often fail to question the validity of their personal beliefs. People manage to find sup­ port for their beliefs, regardless of whether the available evidence is belief con­ sistent, belief inconsistent, or ambiguous. Consequently, previously formed beliefs tend to persist even in the face of contradictory evidence (the perseverance ef­ fect; see Lord, Lepper, & Preston, 1984; Lord, Ross, & Lepper, 1979). Information with extremely negative evaluative implications also is perceived as highly diagnostic and is weighed heavily in judgment, regardless of whether this information is presented vividly or pallidly (the negativity effect; Mizerski, 1982; Skowronski & Carlston, 1987, 1989; Wright, 1974). High-, medium-, and low-quality products often possess many positive features, whereas only lowquality products possess extremely negative ones. Thus, extremely negative in­ formation implies one category (i.e., low quality) to the exclusion of all others. Consequently, the vividness effect is reduced or eliminated when either previ­ ously formed beliefs (the perseverance effect) or extremely negative information (the negativity effect) is available for judgment (Herr et al., 1991). These find­ ings suggest that the accessibility-diagnosticity model provides a useful integra­ tive framework for explaining three seemingly unrelated judgmental phenomena: the vividness effect, the perseverance effect, and the negativity effect. They also explain why the vividness effect has been so elusive in previous research (Taylor & Thompson, 1982). That is, the effects of vividly presented information on judg­ ment are reduced or eliminated when more diagnostic information is available. The accessibility-diagnosticity model also has important implications for un­ derstanding the judgmental effects of measurement operations, rating scales, and survey instruments (Feldman & Lynch, 1988). When product opinions are solicited from consumers, some opinions are retrieved directly from memory, and other opinions are generated or constructed to respond to a judgmental inquiry. Researchers frequently ask consumers for opinions about products or issues they have not really thought about before. However, consumers are surprisingly reluc­ tant to provide a “don’t know” response. Instead, they often prefer to make up a response to answer the question. Feldman and Lynch (1988) referred to this

442

KARDES

type of response as a measurement-induced judgment or self-generated validity. This type of response seriously compromises the conclusions one can draw on the basis of one’s research, because a measurement-induced judgment has no eco­ logical validity (i.e., it would not have been formed in the absence of the research­ er’s intrusive probes). Moreover, a measurement-induced judgment is likely to influence other related judgments, which then influence still other judgments, resulting in a snowball effect. As a consequence, the intercorrelations between beliefs, attitudes, intentions, and behavior that typically are observed in consumer surveys tend to be artificially high (i.e., higher than what would have been ob­ served with less intrusive measures). The development of measurement proce­ dures that control for self-generated validity is an important direction for future research. Several potentially fruitful methodologies should be explored, includ­ ing response-latency procedures (Kardes, 1988b; Stayman & Kardes, 1992) in­ terpolated task procedures (Lichtenstein & Srull, 1985, 1987), and procedures for inferring cognitive representation (Srull, 1984; Wyer & Srull, 1989). The Hypothesis-Testing Model

Hoch and Deighton’s (1989) four-stage model of consumer learning suggests that beliefs about products and services are formed through an iterative process of hypothesis generation and hypothesis testing. Beliefs are viewed as tentatively held hypotheses about the relation between any two concepts, such as price and quality, a brand and an attribute, or an attribute and a benefit. The first stage of the four-stage model is hypothesis generation, because hypotheses must be formed before they can be tested. The remaining three stages pertain to hypothe­ sis testing: exposure to evidence, encoding of evidence, and integration of evi­ dence and prior beliefs. These processes result in revised beliefs. Each of the four stages is influenced by consumer familiarity (i.e., context-specific prior ex­ perience), consumer motivation (i.e., processing goals, involvement, direction, and intensity of cognitive effort), and by information ambiguity (i.e., the extent to which information implies more than one conclusion). Because hypothesis generation requires cognitive effort, consumers are likely to generate hypotheses only when it is functional for them to do so. Consumers who are satisfied with the products and services they use currently are unlikely to consider alternatives, even if better options are available. Expected outcomes and familiar events raise few questions (e.g., why do things seem normal today?) and inhibit hypothesis generation. When consumers are unmotivated to generate their own hypotheses, they passively adopt hypotheses offered by advertisers and test these hypotheses through product trial (Deighton, 1984; Ha & Hoch, 1989; Hoch & Ha, 1986). However, direct behavioral experience with a product often yields ambiguous, inconclusive evidence that tends to be interpreted as support­ ive of the advertiser’s claims. Other possible interpretations are neglected easily (out of sight, out of mind), resulting in overconfident beliefs in the advertisers’

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

4 43

claims. However, under some circumstances, consumers are likely to generate their own hypotheses, especially when consumers fail to achieve their goals (i.e., when a product operates improperly) or when unexpected or novel events are encountered. Even under these circumstances, few hypotheses are generated and alternative hypotheses are overlooked (Fischhoff et al., 1978; Hirt & Castellan, 1988). To test their hypotheses, consumers must gather information either through passive observation or through active self-directed search. Extensive research on multiple cue probability learning has shown that subjects often are unable to learn the relationship between two or more probabilistic cues, especially as error or the number of cues presented increases (Brehmer, 1980). Active search is governed by rules that tend to lend support to the hypothesis under consideration via selec­ tive exposure. For example, the positive-test strategy involves focusing on cases expected to possess (vs. lack) the property of interest (Klayman & Ha, 1987). The good alternative strategy involves focusing on apparently high- (vs. low-) quality alternatives (Meyer, 1987). The single sufficient cause strategy involves focusing on a single possible cause of an event, even when the event is deter­ mined by multiple sufficient causes (Shaklee & Fischhoff, 1982). Each of these strategies involves focusing on some information while neglecting other relevant information. Simplifying strategies can lead to either appropriate or inappropri­ ate conclusions, depending on the interaction between the hypothesized rule and the correct rule (Klayman & Ha, 1987). Hoch and Deighton (1989) maintained that managers deliberately create am­ biguous and difficult learning environments for consumers. Retailers try to avoid carrying the same brands as their competitors. Private label brands, special model numbers of national brands, and different product assortments impede compari­ son shopping across retail outlets. Advertisements often provide information about a single brand to discourage brand comparison, and segmentation strategies often guard against the possibility of leaking product information across different seg­ ments. Negative feedback often is delayed (e.g., it may take years to discover that a kitchen appliance, electric appliance, or automobile is a lemon), and ac­ tions and outcomes often are confounded (e.g., did I break it or did it fall apart?). Selective exposure to incomplete information makes learning difficult, and selective encoding increases learning difficulty even further. Ambiguous evidence often is interpreted as consistent with prior expectations (Darley & Gross, 1983; Rothbart, Evans, & Fulero, 1979; Srull, Lichtenstein, & Rothbart, 1985) or with accessible categories or constructs (Herr, Sherman, & Fazio, 1983; Higgins et al., 1977; Srull & Wyer, 1979, 1980). Because information about products and services is frequently ambiguous and open to multiple interpretation (Hoch & Deighton, 1989), consumers often find that expectancy-guided or primed interpre­ tations fit the data quite readily. When one interpretation fits, consumers often are reluctant to consider other possible interpretations, even though other interpre­ tations fit just as well. Finally, once a decision is made, memory is greater for

444

KARDES

decision-consistent (vs. decision-inconsistent) information (Dellarosa & Bourne, 1984). Differential memory for consistent versus inconsistent information ensures that beliefs formed on the basis of selective exposure and selective encoding will be resistant to change. During the integration stage, consumers combine new information with prior beliefs. New information that supports a prior belief should increase confidence in the belief, whereas new information that contradicts a prior belief should decrease confidence in the belief. However, supportive evidence is often weighed more heavily than contradictory evidence. Information that supports a strongly held prior belief is rated as compelling and probative, whereas information that contradicts a strongly held prior belief is considered weak, flawed, and unin­ formative (Lord et al., 1979, 1984). Similarly, events consistent with predictions based on prior beliefs provide strong support for these beliefs (Gilovich, 1983). However, disconfirmed predictions can easily be explained away or discounted due to bad luck. Finally, people often evaluate the likelihood of observing a par­ ticular datum (D) given that a hypothesis (H) is true. However, they frequently overlook the likelihood of observing D given that the hypothesis is false (H '). Of course, D is uninformative with respect to H if the probability of observing D is the same, regardless of whether H is true or false. For example, the obser­ vation that “Folger’s is mountain grown” (D) supports the hypothesis that “Folger’s is great coffee” (H). However, if all coffee is mountain grown, D supports H ' as well as it supports H, and P(D/H') = P(D/H) = 1. Pseudodiagnosticity, or neglecting the denominator of the likelihood ratio (Do­ herty, Mynatt, Tweney, & Schiavo, 1979; Fischhoff & Beyth-Marom, 1983), encourages decision makers to perceive presented information as relevant, regard­ less of its actual relevance. Consequently, irrelevant information can reduce the judgmental impact of relevant information and prompt decision makers to gener­ ate moderate predictions (the dilution effect; Nisbett, Zukier, & Lemley, 1981). Moreover, insensitivity to alternatives to a focal hypothesis induces decision mak­ ers to ask questions that elicit hypothesis-confirming (vs. discontinuing) answers (Snyder & Swann, 1978; for an alternative interpretation, see Fischhoff & BeythMarom, 1983). Similarly, failing to consider alternatives to a focal hypothesis can lead people to neglect searching for discontinuing evidence (Wason, 1960). In each of these cases, people favor a focal hypothesis and neglect evidence bearing on alternatives to it. When people fail to discriminate between relevant and ir­ relevant information, learning becomes extremely difficult. Although Hoch and Deighton’s (1989) hypothesis-testing model has interest­ ing implications for a wide range of judgmental phenomena, so far the model has been applied only to transformational advertising (Deighton, 1984; Ha & Hoch, 1989; Hoch & Ha, 1986). Product consumption typically provides rich and com­ plex information to consumers. This information often is so rich that nonexperts experience difficulty in deciding on what dimensions to focus. Advertising can guide consumers and transform the consumption experience. For example, novice

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

445

wine drinkers experience difficulty in interpreting the complex flavors they en­ counter when they try a new wine. However, if an ad says a wine has an oaky flavor, it is usually easy to identify the oak. If an ad says a wine has a dry but fruity flavor, these flavors seem to stand out. Transformational advertising makes wines taste richer, clothes look sharper, and cars ride smoother. Hypothesis con­ firming search, encoding, and integration processes can make even the most am­ biguous evidence seem informative. Behavioral Decision Models

One of the main themes of recent behavioral decision research has been that prefer­ ences often are constructed, rather than retrieved, when decision makers need to choose one alternative from a set of alternative products, services, or courses of action (Fischhoff, 1991; Payne, Bettman, & Johnson, 1992; Tversky, Sattah, & Slovic, 1988). Decision makers’ goals, the available alternatives, and the amount of information known about each alternative often vary dramatically across situ­ ations. Consequently, a decision that is effective in one situation may be quite ineffective in another, and decision makers often are forced to treat each situa­ tion as if it were unique. Consumers have a large repertoire of choice heuristics (e.g., conjunctive, disjunctive, lexicographic, and elimination-by-aspects heuris­ tics; Bettman, 1979; Bettman et al., 1991) available for constructing preferences, and the selection of a particular rule or heuristic is contingent on consumers’ goals, the available alternatives, and the amount of information known about each al­ ternative. Another main theme of decision research has been the loss aversion principle, which applies to both risky (i.e., prospect theory; Kahneman & Tversky, 1979, 1984; Tversky & Kahneman, 1981) and riskless choice (i.e., reference depend­ ence theory; Simonson & Tversky, 1992; Tversky & Kahneman, 1991). The loss aversion principle suggests that the value function is steeper for outcomes below a neutral reference point than for outcomes above it, so that “losses loom larger than corresponding gains” (Tversky & Kahneman, 1991, p. 1039). For instance, people typically experience greater displeasure from losing $100 than pleasure from winning $100. This principle potentially explains a surprisingly diverse set of choice phenomena. Examples of these phenomena follow. The Framing Effect. Expected utility theory, the classic normative model of decision making, suggests that consumers confronted with a set of choice op­ tions should select the alternative that maximizes expected utility or offers the most favorable expected outcome (von Neumann & Morgenstern, 1944). This alternative should be selected, regardless of how the choice options are described or framed (the description invariance principle). However, empirical research has shown that preferences differ systematically as a function of the manner in which choice options are described or framed (the framing effect; Tversky &

446

KARDES

Kahneman, 1981). When subjects were asked to choose between a program that would save 200 out of 600 lives for sure (the sure thing option) and a program that would save 600 lives with p = .33 or 0 with p = .67 (the risky option), the majority of subjects (72%) chose the sure thing. However, when an inde­ pendent sample of subjects was asked to choose between a program in which 400 out of 600 people would die for sure (the sure thing option) and a program in which 0 would die with p = .33 or 600 would die with p = .67 (the risky op­ tion), the majority (78%) chose the risky option. Although the outcomes were identical for the sure thing options in both problems, and the outcomes were iden­ tical for the risky options in both problems, subjects chose the sure thing when the outcomes were framed in terms of gains (lives saved), and subjects chose the risky option when outcomes were framed in terms of losses (lives lost). Loss­ es are so much more impactful than corresponding gains that subjects are willing to accept greater risks to avoid losses. Moreover, losses are so much more unat­ tractive than gains are attractive that consumers prefer cash discounts to credit card surcharges (Thaler, 1980) and 75% lean ground beef to 25% fat ground beef (Levin & Gaeth, 1988). The Endowment Effect. Another consequence of loss aversion is that the dis­ pleasure associated with giving up a possession or a privilege is greater than the pleasure associated with attaining the possession or privilege (the endowment ef­ fect; Kahneman et al., 1990; Thaler, 1980). Although mere ownership should not affect the perceived value of an object, loss aversion ensures that buyers and sellers will not see eye to eye. To test this hypothesis, Kahneman et al. (1990) gave some subjects a decorated mug (with a retail value of $5) and later gave them the option to sell the mug. Other subjects received no mug, but later were given the option to buy a mug. Because selling involves losing a mug and buying involves gaining a mug, the median selling price was $7.00 and the median buy­ ing price was $3.50 in one experiment. Similar results were obtained with other objects. The Status Quo Bias. This phenomenon refers to a preference to maintain one’s prior decisions, rather than making new ones (Ritov & Baron, 1992; Samuelson & Zeckhauser, 1988). One’s current state serves as a reference point for evalu­ ating other options, and any change in this state involves loss on some dimensions and gains on other dimensions. Because losses are weighed more heavily than gains, change is often unattractive. Consequently, people prefer to keep their cur­ rent jobs, automobile color, financial investments, and medical insurance poli­ cies (Samuelson & Zeckhauser, 1988). Tradeoff Contrast. Just as a circle seems large when surrounded by small circles and small when surrounded by large circles (contrast effects), a product seems attractive when compared with less attractive products and unattractive

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

447

when compared with more attractive products (the attraction effect; Huber, Payne, & Puto, 1982; Simonson, 1989). Simonson and Tversky (1992) recently extend­ ed this notion to tradeoffs between attributes (tradeoff contrast). For example, many purchase decisions involve a tradeoff between price and quality. Suppose a consumer is asked to choose between a $1200 personal computer with 960K (option A) and a $1000 personal computer with 640K (option B). Tradeoff con­ trast implies that option A (which involves a tradeoff of $200 for an additional 320K) will be preferred more if the consumer considers (vs. fails to consider) a third option of a $1300 personal computer with 960K (which involves a tradeoff of $300 for an additional 320K). Simonson and Tversky found support for the tradeoff contrast principle across a wide range of products. Extremeness Aversion. Suppose a customer compares a brand high in both price and quality (option A) with a brand that is moderate in both price and quali­ ty (option B). Option B will be preferred more if the consumer also considers a third option, C, that is low in both price and quality (the compromise effect; Simonson, 1989). Because disadvantages loom larger than corresponding advan­ tages, the intermediate option B is preferred, because it offers small disadvan­ tages on both dimensions, whereas option A offers a large price disadvantage and option C offers a large quality disadvantage (extremeness aversion; Simon­ son & Tversky, 1992). Simonson and Tversky (1992) found support for the ex­ tremeness aversion principle across a wide variety of products. The Sunk Cost Fallacy. Economists maintain that only incremental costs and benefits should affect decisions, and that historical costs should be irrelevant. However, after investing a lot of time, money, or effort in a project, people often feel reluctant to abandon the project even if they would be better off doing so (the sunk cost fallacy; Arkes & Blumer, 1985; Larrick, Morgan, & Nisbett, 1990; Thaler, 1980). For example, many steel companies went out of business, because they continued to rely on outdated mixing technologies even when superior tech­ nologies became available. After investing millions of dollars in old technolo­ gies, abandoning these technologies seemed like abandoning millions of dollars. Similarly, after spending over one billion dollars on the Tennessee-Tombigbee Waterway Project, politicians were reluctant to abandon the project, even though it was clear that the finished project would be worth less than the amount of money required to complete it (Arkes & Blumer, 1985). The loss aversion principle sug­ gests that people are reluctant to give up what they already have invested, even when the investment is unlikely to ever pay off. Regret. Regret refers to feelings of loss or sorrow that follow when a deci­ sion turns out badly. Feelings of regret are especially intense when negative out­ comes result from action as opposed to inaction (the omission bias; Ritov & Baron, 1992) or from atypical as opposed to typical or normal action (Kahneman & Miller,

448

KARDES

1986). Concern about the regret that may follow a bad decision promotes ex­ treme risk aversion. Sometimes risk aversion is so extreme that a decision maker is reluctant to make a decision (Thaler, 1980). Moreover, concern about regret induces consumers to purchase a product on sale now, rather than wait for a bet­ ter sale later, and induces consumers to prefer a well-known brand to a less ex­ pensive, lesser known brand (Simonson, 1992). In each of these cases, consumers are willing to pay a premium to reduce uncertainty and to reduce the risk of ob­ taining an unsatisfactory outcome. The MODE Model

This chapter reviewed several models of judgment and highlighted the factors emphasized by each model. These models become more useful when one con­ siders an additional model bearing on the translation of judgment into overt be­ havior (i.e., choices and purchase decisions). Fazio’s (1990) MODE model integrates a wide range of processes by which attitudinal judgments (overall evalu­ ations) can be translated into overt behavior. Before describing the MODE model, it is useful to review two extreme processes that lie at opposite ends of the spontaneous-deliberative processing continuum. Many everyday decisions are made quickly and easily, because they are based on a spontaneous attitude-to-behavior process that does not require a great deal of deliberation or reflection. If every decision required extensive analysis and effortful information processing, time and energy constraints would allow peo­ ple to make very few decisions each day and it would take much longer for peo­ ple to achieve their goals. Fortunately, people form attitudes that summarize large amounts of information, reduce cognitive load, and simplify decision making. Fazio’s (1986, 1989) spontaneous processing model of the attitude-behavior re­ lation suggests that attitudes must be activated before they can influence behavior, and the likelihood of attitude activation on mere observation of the attitude ob­ ject depends on the accessibility of the attitude or the strength of the objectevaluation association. Attitude accessibility (the readiness with which an atti­ tude is retrieved from memory) is operationalized in terms of latency of response to an attitudinal inquiry. Any manipulation of the strength of the object-evaluation association—including manipulations such as repetitive advertising (Berger & Mitchell, 1989), repeated attitude activation (Fazio, Sanbonmatsu, Powell, & Kardes, 1986), attitude formation based on direct behavioral experience (Fazio, Chen, McDonel, & Sherman, 1982), or effortful cognitive processing (Kardes, 1988b; Stayman & Kardes, 1992)—results in a corresponding increase in the speed with which subjects can respond to attitudinal inquiries. Once activated, an atti­ tude directs focal attention toward attitude-consistent aspects of an attitude ob­ ject and leads consumers to neglect attitude-inconsistent information. Hence, a positive attitude leads consumers to focus on positive aspects of a product, whereas a negative attitude leads consumers to focus on negative aspects of a product.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

449

Selective attention induces consumers to perceive an attitude object as more con­ sistent with an attitude than it really is (an assimilation effect), and this percep­ tion influences the definition of the situation (e.g., it is time to buy or consume the product, or it is time to reject or discard the product). Of course, norms also impinge on the definition of the situation (you cannot buy the product if you lack cash or credit). The definition of the situation then serves as the direct deter­ minant of overt behavior. Fazio, Powell, and Williams (1989) tested the spontaneous processing model in a consumer setting by asking subjects to indicate their attitudes toward a large number of products (e.g., Sun Maid raisins, Dentyne gum, Mounds candy) by pressing either a “like” or a “dislike” button (speed and accuracy was emphasized). On the basis of these response latencies, each of 10 target products were catego­ rized into high, medium, or low attitude-accessibility groups for each subject. Subjects then rated these products on standard 7-point attitude scales. Finally, at the end of the experiment, the 10 target products were presented on a table and subjects were permitted to choose 5 products to take home. As predicted, across subjects and across the 10 products, as attitude accessibility increased, attitude-behavior consistency also increased. Moreover, low attitude-accessibility products were more likely to be selected when they were displayed prominently in the front row as opposed to less prominently in the back row (the 10 products were displayed in two rows of 5). Hence, when attitude accessibility was low and, therefore, when attitudes were unlikely to influence choice, the manner in which the products were displayed had a strong impact on choice decisions. Of course, not all choice decisions are made quickly and easily. Some deci­ sions require much deliberation and reflection. The theory of reasoned action (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975; Sheppard, Hartwick, & Warshaw, 1988) provides a good example of a deliberative processing model of the attitude-behavior relation. This theory suggests that attitudes are formed on the basis of a careful analysis of the costs and benefits associated with performing a specific action (e.g., purchasing brand A). Attitude toward the behavior then influences consumers’ intentions. Intentions also are influenced by subjective norms or beliefs about what significant others will think of one if one performs the be­ havior in question. Intentions then serve as the direct and immediate determinant of overt behavior. Fazio’s (1990) MODE model provides a useful framework for integrating the spontaneous (Fazio, 1986, 1989) and the deliberative (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975) attitude-to-behavior process models. The MODE model suggests that Motivation and Opportunity serve as Determinants of the process by which attitudes guide behavior. The model can be viewed as a specific appli­ cation of Kruglanski’s (1989) theory of lay epistemology, which suggests that need for structure (a preference for a conclusion, any conclusion, as opposed to ambiguity and confusion) and fear of invalidity (a preference for a valid con­ clusion, because error could lead to aversive consequences) jointly influence

450

KARDES

processing effort. In the present context, this theory predicts that effortful, deliber­ ative, data-driven attitude-to-behavior processes are most likely when both need for structure is low and fear of invalidity is high. By contrast, spontaneous, theorydriven attitude-to-behavior processes are likely when either need for structure is high or fear of invalidity is low. To test the MODE model, Sanbonmatsu and Fazio (1990) asked subjects to form overall evaluations of two department stores, Smith’s and Brown’s, while view­ ing information about the specific departments (e.g., clothing, jewelry) of each store. The majority of Smith’s departments were described very positively, whereas the majority of Brown’s departments were described very negatively. Consequent­ ly, subjects formed much more favorable overall evaluations of Smith’s than of Brown’s. However, the camera departments of both stores were described posi­ tively, and the Brown’s camera department was described more positively than Smith’s. Later, subjects were asked to imagine that they needed to buy a camera, and to indicate which store they should visit. In high need for structure conditions, subjects were allowed 15 seconds to make a decision. In low need for structure conditions, no time limit was imposed. In high fear of invalidity conditions, sub­ jects were told that their decisions would be compared with the decisions of other subjects and that they would have to explain their decisions to the other subjects and the experimenter. These instructions were omitted in low fear of invalidity conditions. As predicted, when need for structure was low and fear of invalidity was high, subjects preferred to shop for cameras at Brown’s (the store with the superior camera department). Conversely, when need for structure was high or when fear of invalidity was low, subjects preferred to shop for cameras at Smith’s (the generally superior store with the inferior camera department). Hence, when effortful, data-driven processing was likely (i.e., when need for structure was low and fear of invalidity was high), subjects searched memory for information about specific departments and selected the store with the better camera depart­ ment. However, when less effortful, theory-driven processing was likely (i.e., when need for structure was high or fear of invalidity was low), subjects selected the store for which they had formed a more favorable overall evaluation. Finally, the spontaneous processing model and the deliberative processing model represent the extremes of the spontaneous-deliberative processing continu­ um. Fazio (1990) suggested that many decisions lie more toward the middle of the continuum, where some aspects of the process are spontaneous and other aspects are deliberative. Future research should attempt to identify mixed processes involved in the translation of judgments into behavior. CONCLUSION Judgments of products and services are influenced by a wide variety of psycho­ logical processes. Selective exposure, selective attention, and selective encoding processes determine what information is likely to be gathered and stored in

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

451

memory. The manner in which information is organized in memory, and the read­ iness with which it can be retrieved, determine what information is likely to be used as an input to judgment. Theories of consumer judgment attempt to identify the key processes involved in construing the implications of information and in weighing and integrating these implications to generate new judgments or to ad­ just previously existing judgments. Although a greater emphasis on theory is need­ ed for developing a deeper understanding of consumer judgment and decision processes, the few existing theories show considerable promise as frameworks for integrating many seemingly unrelated judgmental phenomena and for gener­ ating hypotheses for future research.

REFERENCES Aaker, D. A., & Keller, K. L. (1990). Consumer evaluations of brand extensions. Journal o f M ar­ keting, 54, 27-41. Aaker, D. A., Stayman, D. M ., & Hagerty, M. R. (1986). Warmth in advertising: Measurement, impact, and sequence effects. Journal o f Consumer Research, 12, 365-381. Agostinelli, G., Sherman, S. J., Fazio, R. H., & Hearst, E. S. (1986). Detecting and identifying change: Additions versus deletions. Journal o f Experimental Psychology: Human Perception and Performance, 12, 445-454. Ajzen, I., & Fishbein, M. (1989). Understanding attitudes and predicting social behavior. Engle­ wood Cliffs, NJ: Prentice-Hall. Alba, J. W. (1984). Nature of inference representation. American Journal o f Psychology, 97, 215-233. Alba, J. W ., & Chattopadhyay, A. (1985). Effects of context and part-category cues on recall of competing brands. Journal o f Marketing Research, 22, 340-349. Alba, J. W ., & Chattopadhyay, A. (1986). Salience effects in brand recall. Journal o f Marketing Research, 23, 363-369. Alba, J. W ., & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal o f Consumer Research, 13, 411-454. Alba, J. W ., Hutchinson, J. W ., & Lynch, J. G. (1991). Memory and decision making. In T. S. Robertson & H. H. Kassarjian (Eds.), Handbook o f consumer behavior (pp. 1-49). Englewood Cliffs, NJ: Prentice-Hall. Allen, C. T ., & Janiszewski, C. (1989). Assessing the role of contingency awareness in attitudinal conditioning with implications for advertising research. Journal o f Marketing Research, 26, 30-43. Allen, C. T ., & Madden, T. J. (1985). A closer look at classical conditioning. Journal o f Consumer Research, 12, 301-315. Anderson, J. R. (1983). The architecture o f cognition. Cambridge, MA: Harvard University Press. Anderson, J. R., & Bower, G. H. (1973). Human associative memory. Hillsdale, NJ: Lawrence Erl­ baum Associates. Anderson, N. H. (1967). Averaging model analysis of set-size effect in impression formation. Jour­ nal o f Experimental Psychology, 75, 158-165. Anderson, N. H. (1970). Functional measurement and psychological judgment. Psychological Review, 77, 153-170. Anderson, N. H. (1981). Foundations o f information integration theory. New York: Academic Press. Anderson, N. H. (1982). Methods o f information integration theory. New York: Academic Press. Anderson, N. H. (1991). Contributions to information integration theory (Vols. I—III). Hillsdale, NJ: Lawrence Erlbaum Associates.

452

KARDES

Anderson, N. H., & Hubert, S. (1963). Effects of concomitant verbal recall on order effects in per­ sonality impression formation. Journal o f Verbal te a m in g and Verbal Behavior, 2, 379-391. Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Organizational Behavior and Human Decision Processes, 35, 129-140. Arkes, H. R., Boehm, L. E., & Xu, G. (1991). Determinants of judged validity. Journal o f Ex­ perim ental Social Psychology, 27, 576-605. Arkes, H. R., & Freedman, M. R. (1984). A demonstration of the costs and benefits of expertise in recognition memory. M emory and Cognition, 12, 84-89. Arkes, H. R., Hackett, C ., & Boehm, (1989). The generality of the relation between familiarity and judged validity. Journal o f Behavioral Decision Making, 2, 81-94. Baker, W. E., & Lutz, R. J. (1988). The relevance-accessibility model of advertising effectiveness. In S. Hecker & D. M. Stewart (Eds.), Nonverbal communications in advertising (pp. 59-84). Lex­ ington, MA: Lexington. Barsalou, L. W. (1985). Ideals, central tendency, and frequency of instantiation as determinants of graded structure. Journal o f Experimental Psychology: te a m in g , Memory, and Cognition, 11, 629-654. Barsalou, L. W ., & Ross, B. H. (1986). The roles of automatic and strategic processing in sensitivi­ ty to superordinate and property frequency. Journal o f Experimental Psychology: tea m in g , Memory, and Cognition, 12, 116-134. Batra, R., & Ray, M. L. (1986). Affective responses mediating acceptance of advertising. Journal o f Consumer Research, 13, 234-249. Beatty, S. E., & Smith, S. M. (1987). External search effort: An investigation across several product categories. Journal o f Consumer Research, 14, 83-95. Beckwith, N. E., & Lehmann, D. R. (1975). The importance of halo effects in multi-attribute atti­ tude models. Journal o f Marketing Research, 12, 265-275. Belmore, S. M. (1981). Imagery and semantic elaboration in hypermnesia for words. Journal o f Ex­ perimental Psychology: Human te a m in g and Memory, 7, 191-203. Bern, D. J. (1972). Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 6, pp. 1-62). New York: Academic Press. Berger, I. E., & Mitchell, A. A. (1989). The effect of advertising on attitude accessibility, attitude confidence, and the attitude-behavior relationship. Journal o f Consumer Research, 16, 269-279. Bettman, J. R. (1979). An information processing theory o f consumer choice. Reading, MA: AddisonWesley. Bettman, J. R. (1986). Consumer psychology. Annual Review o f Psychology, 37, 257-289. Bettman, J. R., Capon, N., & Lutz, R. J. (1975). Multiattribute measurement models and multiat­ tribute theory: A test of construct validity. Journal o f Consumer Research, 1, 1-15. Bettman, J. R., Johnson, E. J., & Payne, J. W. (1991). Consumer decision making. In T. S. Robert­ son & H. H. Kassarjian (Eds.), Handbook o f consumer behavior (pp. 50-83). Englewood Cliffs, NJ: Prentice-Hall. Bettman, J. R., & Park, C. W. (1980). Effects of prior knowledge, experience, and phase of the choice process on consumer decision processes: A protocol analysis. Journal o f Consumer Research, 7, 234-248. Bettman, J. R., & Sujan, M. (1987). Effects of framing on evaluation of comparable and noncom­ parable alternatives by expert and novice consumers. Journal o f Consumer Research, 14, 141-154. Biehal, G. J., & Chakravarti, D. (1982). Information presentation format and learning goals as de­ terminants of consumers’ memory retrieval and choice processes. Journal o f Consumer Research, 8, 431-441. Biehal, G. J., & Chakravarti, D. (1983). Information accessibility as a moderator of consumer choice. Journal o f Consumer Research, 10, 1-14. Biehal, G. J., & Chakravarti, D. (1986). Consumers’ use of memory and external information in choice: Macro and micro processing perspectives. Journal o f Consumer Research, 12, 382-405.

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

45 3

Bierley, C ., McSweeney, F. K., & Vannieuwkerk, R. (1985). Classical conditioning of preferences for stimuli. Journal o f Consumer Research, 12, 316-323. Bodenhausen, G. V., & Lichtenstein, M. (1987). Social stereotypes and information processing strate­ gies: The impact of task complexity. Journal o f Personality and Social Psychology, 52, 871-880. Bodenhausen, G. V., & Wyer, R. S. (1985). Effects of stereotypes on decision-making and informationprocessing strategies. Journal o f Personality and Social Psychology, 48, 267-282. Borgida, E., & Howard-Piney, B. (1983). Personal involvement and the robustness of perceptual salience effects. Journal o f Personality and Social Psychology, 45, 560-570. Brehmer, B. (1980). In one word: Not from experience. Acta Psychologica, 45, 223-241. Brewer, M. (1988). A dual process model of impression formation. In R. S. Wyer & T. K. Srull (Eds.), Advances in social cognition: Vol. 1. A dual-process model o f impression form ation (pp. 1-36). Hillsdale, NJ: Lawrence Erlbaum Associates. Britt, S. H., Adams, S. C ., & Miller, A. S. (1972). How many advertising exposures per day? Jour­ nal o f Advertising Research, 12, 3-9. Brown, S. P., & Stayman, D. M. (1992). Antecedents and consequences of attitude toward the ad: A meta-analysis. Journal o f Consumer Research, 19, 34-51. Bracks, M. (1985). The effects of product class knowledge on information search behavior. Journal o f Consumer Research, 12, 1-16. Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64, 123-152. Burke, R. R., & Srull, T. K. (1988). Competitive interference and consumer memory for advertis­ ing. Journal o f Consumer Research, 15, 55-68. Cacioppo, J. T ., & Petty, R. E. (1982). The need for cognition. Journal o f Personality and Social Psychology, 42, 116-131. Cacioppo, J. T., & Petty, R. E. (1985). Central and peripheral routes to persuasion: The role of message repetition. In L. F. Alwitt & A. A. Mitchell (Eds.), Psychological processes and adver­ tising effects: Theory, research, and application (pp. 91-111). Hillsdale, NJ: Lawrence Erlbaum Associates. Calder, B. J., & Sternthal, B. (1980). Television commercial wearout: An information processing view. Journal o f Marketing Research, 17, 173-186. Cann, A., Sherman, S. J., & Elkes, R. (1975). Effects of initial request size and timing of a second request on compliance : The foot-in-the-door and door-in-the-face. Journal o f Personality and So­ cial Psychology, 32, 774-782. Carlston, D. E. (1980). The recall and use of traits and events in social inference processes. Journal o f Experimental Social Psychology, 16, 303-328. Carlston, D. E., & Skowronski, J. J. (1986). Trait memory and behavior memory: The effects of alternative pathways on impression judgment response times. Journal o f Personality and. Social Psychology, 50, 5-13. Carpenter, G. S., & Nakamoto, K. (1989). Consumer preference formation and pioneering advan­ tage. Journal o f Marketing Research, 26, 285-298. Cattin, P., & Wittink, D. R. (1982). Commercial use of conjoint analysis: A survey. Journal o f Mar­ keting, 46, 44-53. Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal o f Personality and Social Psychology, 39, 752-766. Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and systematic information processing within and beyond the persuasion context. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 212-252). New York: Guilford. Chaiken, S., & Stangor, C. (1987). Attitudes and attitude change. Annual Review o f Psychology, 38, 575-630. Chanowitz, B., & Langer, E. J. (1981). Premature cognitive commitment. Journal o f Personality and Social Psychology, 41, 1051-1063. Chattopadhyay, A., & Alba, J. W. (1988). The situational importance of recall and inference in con­ sumer decision making. Journal o f Consumer Research, 15, 1-12.

454

KARDES

Childers, T. L., & Houston, M. J. (1984). Conditions for a picture-superiority effect on consumer theory. Journal o f Consumer Research, 15, 643-654. Childers, T. L ., Houston, M. J., & Heckler, S. E. (1985). Measurement of individual differences in visual versus verbal information processing. Journal o f Consumer Research, 12, 125-134. Cialdini, R. B. (1988). Influence: Science and practice. Glenview, IL: Scott, Foresman. Cialdini, R. B., Levy, A ., Herman, C. P., & Evenbeck, S. (1973). Attitudinal politics: The strategy of moderation. Journal o f Personality and Social Psychology, 25, 100-108. Cialdini, R. B., Vincent, J. E., Lewis, S. K., Catalan, J., Wheeler, D ., & Darby, B. L. (1975). Reciprocal concessions procedure for inducing compliance: The door-in-the-face technique. Jour­ nal o f Personality and Social Psychology, 31, 206-215. Cohen, J. B., & Chakravarti, D. (1990). Consumer psychology. Annual Review o f Psychology, 41, 243-288. Collins, A. M ., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psy­ chological Review, 82, 407-428. Corbett, A. T ., & Dosher, B. A. (1978). Instrument inferences in sentence encoding. Journal o f Verbal Learning and Verbal Behavior, 17, 479-491. Darley, J. M ., & Gross, P. H. (1983). A hypothesis-confirming bias in labelling effects. Journal o f Personality and Social Psychology, 44, 20-33. Dawes, R. M. (1988). Rational choice in an uncertain world. San Diego, CA: Harcourt Brace Jovanovich. Deighton, J. (1984). The interaction of advertising and evidence. Journal o f Consumer Research, 11, 763-770. Dellarosa, D., & Bourne, L. E. (1984). Decisions and memory: Differential retrievability of consis­ tent and contradictory evidence. Journal o f Verbal Learning and Verbal Behavior, 23, 669-682. Dick, A., Chakravarti, D., & Biehal, G. (1990). Memory-based inferences during choice. Journal o f Consumer Research, 17, 82-93. Doherty, M. E., Mynatt, C. R., Tweney, R. D., & Schiavo, M. D. (1979). Pseudodiagnosticity. Acta Psychologica, 43, 111-121. Duncan, C. P., & Olshavsky, R. W. (1982). External search: The role of consumer beliefs. Journal o f Marketing Research, 19, 32-43. Erdelyi, M. H ., & Appelbaum, G. A. (1973). Cognitive masking: The disruptive effects of an emo­ tional stimulus upon the perception of contiguous neutral items. Bulletin o f the Psychonomic Soci­ ety, 1, 56-61. Erdelyi, M., & Kleinbard, J. (1978). Has Ebbinghaus decayed with time? The growth of recall (hypermnesia) over days. Journal o f Experimental Psychology: Human Learning and Memory, 4, 275-289. Ericsson, K. A., & Simon, H. A. (1980). Verbal reports as data. Psychological Review, 87, 215-251. Farquhar, P. H. (1989). Managing brand equity. Marketing Research, 1, 24-33. Farquhar, P. H., Herr, P. M ., & Fazio, R. H. (1990). A relational model for category extensions of brands. Advances in Consumer Research, 17, 856-860. Fazio, R. H. (1981). On the self-perception explanation of the overjustification effect: The role of salience of initial attitude. Journal o f Experimental Social Psychology, 17, 417-426. Fazio, R. H. (1986). How do attitudes guide behavior? In R. M. Sorrentino & E. T. Higgins (Eds.), The handbook o f motivation and cognition: Foundations o f social behavior (pp. 204-243). New York: Guilford. Fazio, R. H. (1989). On the power and functionality of attitudes. The role of attitude accessibility. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 153-179). Hillsdale, NJ: Lawrence Erlbaum Associates. Fazio, R. H. (1990). Multiple processes by which attitudes guide behavior: The MODE model as an integrative framework. In M. P. Zanna (Ed.), Advances in experimental social psychology (pp. 75-109). New York: Academic Press. Fazio, R. H., Chen, J., McDonel, E. C., & Sherman, S. J. (1982). Attitude accessibility, attitudebehavior consistency, and the strength of the object-evaluation association. Journal o f Experimen­ tal Social Psychology, 18, 339-357.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

45 5

Fazio, R. H., Herr, P. M ., & Powell, M. C. (1992). On the development and strength of categorybased associations in memory: The case of mystery ads. Journal o f Consumer Psychology, 7, 1-14. Fazio, R. H., Powell, M. C., & Williams, C. J. (1989). The role of attitude accessibility in the attitudeto-behavior process. Journal o f Consumer Research, 16, 280-288. Fazio, R. H., Sanbonmatsu, D. M ., Powell, M. C ., & Kardes, F. R. (1986). On the automatic acti­ vation of attitudes. Journal o f Personality and Social Psychology, 50, 229-238. Fazio, R. H., & Zanna, M. P. (1981). Direct experience and attitude-behavior consistency. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 14, pp. 162-202). New York: Academic Press. Feldman, J. M ., & Lynch, J. G. (1988). Self-generated validity and other effects of measurement on belief, attitude, intention, and behavior. Journal o f A pplied Psychology, 73, 421-435. Fischhoff, B. (1991). Value elicitation: Is there anything in there? American Psychologist, 46, 835-847. Fischhoff, B., & Beyth-Marom, R. (1983). Hypothesis evaluation from a Bayesian perspective. P sy­ chological Review, 90, 239-260. Fischhoff, B., Slovic, P., & Lichtenstein, S. (1978). Fault trees: Sensitivity of estimated failure prob­ abilities to problem representation. Journal o f Experimental Psychology: Human Perception and Performance, 4, 330-344. Fishbein, M ., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley. Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression formation, from category-based to individuating processes: Influences of information and motivation on attention and interpreta­ tion. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 1-74). New York: Academic Press. Fiske, S. T., & Pavelchak, M. A. (1986). Category-based versus piecemeal-based affective responses: Developments in schema-triggered affect. In R. M. Sorrentino & E. T. Higgins (Eds.), The hand­ book o f motivation and cognition: Foundations o f social behavior (pp. 167-203). New York: Guilford. Fiske, S. T ., & Taylor, S. E. (1991). Social cognition. New York: McGraw-Hill. Ford, G. T ., & Smith, R. A. (1987). Inferential beliefs in consumer evaluations: An assessment of alternative processing strategies. Journal o f Consumer Research, 14, 363-371. Freedman, J., & Fraser, S. (1966). Compliance without pressure: The foot-in-the-door technique. Journal o f Personality and Social Psychology, 4, 195-202. Friedman, A., & Poison, M. C. (1981). Hemispheres as independent resource systems: Limitedcapacity processing and cerebral specialization. Journal o f Experimental Psychology: Human P er­ ception and Performance, 7, 1031-1058. Furse, D. H., Punj, G. N., & Stewart, D. W. (1984). A typology of individual search strategies among purchasers of new automobiles. Journal o f Consumer Research, 10, 417-431. Gavanski, I., & Hui, C. (1992). Natural sample spaces and uncertain belief. Journal o f Personality and Social Psychology, 63, 766-780. Gilbert, D. T., Pelham, B. W ., & Krull, D. S. (1988). On cognitive busyness: When person per­ ceivers meet person’s perceived. Journal o f Personality and Social Psychology, 54, 733-739. Gilovich, T. (1983). Biased evaluation and persistence in gambling. Journal o f Personality and So­ cial Psychology, 44, 1110-1126. Gleicher, F. H., Kost, K. A., Baker, S. M .r Strathman, A., Richman, S. A., & Sherman, S. J. (1990). The role of counterfactual thinking in judgments of affect. Personality and Social Psychol­ ogy Bulletin, 16, 284-295. Gorn, G. J. (1982). The effects of music in advertising on choice behavior: A classical conditioning approach. Journal o f Marketing, 46, 94-101. Gotz, A., & Jacoby, L. L. (1974). Encoding and retrieval processes in long-term retention. Journal o f Experimental Psychology, 102, 291-297. Greenwald, A. G. (1968). Cognitive learning, cognitive response to persuasion, and attitude change. In A. G. Greenwald, T. C. Brock, & T. M. Ostrom (Eds.), Psychological foundations o f attitudes (pp. 147-170). New York: Academic Press.

456

KARDES

Greenwald, A. G. (1988). Self-knowledge and self-deception. In J. S. Lockard & D. L. Paulhus (Eds.), Self-deception: An adaptive mechanism ? (pp. 113-131). Englewood Cliffs, NJ: Prentice-Hall. Greenwald, A. G., & Leavitt, C. (1984). Audience involvement in advertising: Four levels. Journal o f Consumer Research, 11, 581-592. Griffin, D. W ., Dunning, D., & Ross, L. (1990). The role of construal processes in overconfident predictions about the self and others. Journal o f Personality and Social Psychology, 59, 1128-1139. Griffin, D. W ., & Tversky, A. (1992). The weighing of evidence and the determinants of confi­ dence. Cognitive Psychology, 24, 411-435. Gurumurthy, K., & Urban, G. (1992). Dynamic effects of the order of entry on market share, trial penetration, and repeat purchases for frequently purchased consumer goods. M arketing Science, 11, 235-250. Ha, Y., & Hoch, S. J. (1989). Ambiguity, processing strategy, and advertising-evidence interac­ tions. Journal o f Consumer Research, 16, 354-360. Hannah, D., & Sternthal, B. (1984). Detecting and explaining the sleeper effect. Journal o f Con­ sumer Research, 11, 632-642. Harris, R. J. (1977). Comprehension of pragmatic implications in advertising. Journal o f Applied Psychology, 62, 603-608. Harris, R. J., & Monaco, G. E. (1978). Psychology of pragmatic implications in advertising: Infor­ mation processing between the lines. Journal o f Experimental Psychology: General, 107, 1-22. Hasher, L., Goldstein, D ., & Toppino, T. (1977). Frequency and the conference of referential valid­ ity. Journal o f Verbal Learning and Verbal Behavior, 16, 107-112. Hastak, M ., & Olson, J. C. (1989). Assessing the role of brand-related cognitive responses as medi­ ators of communication effects on cognitive structure. Journal o f Consumer Research, 15, 444-456. Hastie, R. (1981). Schematic principles in human memory. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario Symposium (Vol. 1, pp. 39-88). Hillsdale, NJ: Lawrence Erlbaum Associates. Hastie, R. (1983). Social inference. Annual Review o f Psychology, 34, 511-542. Hastie, R. (1984). Causes and effects of causal attribution. Journal o f Personality and Social P sy­ chology, 46, 44-56. Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the judgment task is memory-based or on-line. Psychological Review, 93, 258-268. Hauser, J. R., & Wemerfelt, B. (1990). An evaluation cost model of consideration sets. Journal o f Consumer Research, 16, 393-408. Hawkins, S. A., & Hoch, S. J. (1992). Low-involvement learning: Memory without evaluation. Journal o f Consumer Research, 19, 212-225. Herr, P. M. (1989). Priming price: Prior knowledge and context effects. Journal o f Consumer Research, 16, 67-75. Herr, P. M ., Kardes, F. R., & Kim, J. (1991). Effects of word-of-mouth and product-attribute infor­ mation on persuasion: An accessibility-diagnosticity perspective. Journal o f Consumer Research, 17 , 454-462. Herr, P. M ., Sherman, S. J., & Fazio, R. H. (1983). On the consequences of priming: Assimilation and contrast effects. Journal o f Experimental Social Psychology, 19, 323-340. Higgins, E. T ., Bargh, J. A., & Lombardi, W. (1985). The nature of priming effects on categoriza­ tion. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 11, 59-69. Higgins, E. T ., Rholes, W. S., & Jones, C. R. (1977). Category accessibility and impression forma­ tion. Journal o f Experimental Social Psychology, 13, 141-154. Hirt, E. R., & Castellan, N. J. (1988). Probability and category redefinition in the fault tree paradigm. Journal o f Experimental Psychology: Human Perception and Performance, 14, 122-131. Hoch, S. J. (1984). Availability and interference in predictive judgment. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 10, 649-662. Hoch, S. J. (1985). Counterfactual reasoning and accuracy in predicting personal events. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 11, 719-731.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

45 7

Hoch, S. J., & Deighton, J. (1989). Managing what consumers learn from experience. Journal o f Marketing, 53, 1-20. Hoch, S. J., & Ha, Y. (1986). Consumer learning: Advertising and the ambiguity of product ex­ perience. Journal o f Consumer Research, 13, 221-233. Hogarth, R. M ., & Einhom, H. J. (1992). Order effects in belief updating: The belief-adjustment model. Cognitive Psychology, 24, 1-55. Holland, J. H., Holyoak, K. J., Nisbett, R. E ., & Thagard, P. T. (1986). Induction: Processes o f inference, learning, and discovery. Cambridge, MA: MIT Press. Hong, S., & Wyer, R. S. (1989). Effects of country-of-origin and product-attribute information on product evaluation: An information processing perspective. Journal o f Consumer Research, 16, 175-187. Hong, S., & Wyer, R. S. (1990). Determinants of product evaluation: Effects of the time interval between knowledge of a product’s country of origin and information about its specific attributes. Journal o f Consumer Research, 17, 277-288. Houston, M. J., Childers, T. L., & Heckler, S. E. (1987). Picture-word consistency and the elaborative processing of advertisements. Journal o f Marketing Research, 24, 359-370. Howard-Pitney, B., Borgida, E ., & Omoto, A. M. (1986). Personal involvement: An examination of processing differences. Social Cognition, 4, 39-57. Hoyer, W. D. (1984). An examination of consumer decision making for a common repeat purchase product. Journal o f Consumer Research, 11, 822-829. Hoyer, W. D., & Brown, S. P. (1990). Effects of brand awareness on choice for a common, repeatpurchase product. Journal o f Consumer Research, 17, 141-148. Huber, J., & McCann, J. W. (1982). The impact of inferential beliefs on product evaluations. Jour­ nal o f Marketing Research, 19, 324-333. Huber, J., Payne, J. W ., & Puto, C. (1982). Adding asymmetrically dominated alternatives: Viola­ tions of regularity and the similarity hypothesis. Journal o f Consumer Research, 9, 90-98. Jaccard, J., & Wood, G. (1988). The effects of incomplete information on the formation of attitudes toward behavioral alternatives. Journal o f Personality and Social Psychology, 54, 580-591. Janiszewski, C. (1988). Preconscious processing effects: The independence of attitude formation and conscious thought. Journal o f Consumer Research, 15, 199-209. Janiszewski, C. (1990). The influence of print advertisement organization on affect toward a brand name. Journal o f Consumer Research, 17, 53-65. Johnson, B. T ., & Eagly, A. H. (1989). The effects of involvement on persuasion: A meta-analysis. Psychological Bulletin, 106, 290-314. Johnson, B. T., & Eagly, A. H. (1990). Involvement and persuasion: Types, traditions, and the evi­ dence. Psychological Bulletin, 107, 375-384. Johnson, E. J., & Russo, J. E. (1984). Product familiarity and learning new information. Journal o f Consumer Research, 11, 542-550. Johnson, R. D., & Levin, I. P. (1985). More than meets the eye: The effect of missing information on purchase evaluations. Journal o f Consumer Research, 12, 169-177. Judd, C. M ., & Kulik, J. A. (1980). Schematic effects of social attitudes on information processing and recall. Journal o f Personality and Social Psychology, 38, 569-578. Kahle, L. R., & Homer, P. M. (1985). Physical attractiveness of the celebrity endorser: A social adaptation perspective. Journal o f Consumer Research, 11, 954-961. Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice-Hall. Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1990). Experimental tests of the endowment effect and the Coase theorem. Journal o f Political Economy, 98, 1325-1348. Kahneman, D., & Miller, D. T. (1986). Norm theory: Comparing reality to its alternatives. Psycho­ logical Review, 93, 136-153. Kahneman, D., & Treisman, A. (1984). Changing views of attention and automaticity. In R. Parasuraman & R. Davies (Eds.), Varieties o f attention (pp. 29-61). New York: Academic Press.

458

KARDES

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica , 47, 263-291. Kahneman, D., & Tversky, A. (1984). Choice, values, and frames. American Psychologist, 39, 341-350. Kardes, F. R. (1986). Effects of initial product judgments on subsequent memory-based judgments. Journal o f Consumer Research, 13, 1-11. Kardes, F. R. (1988a). A nonreactive measure of inferential beliefs. Psychology and Marketing, 5, 273-286. Kardes, F. R. (1988b). Spontaneous inference processes in advertising: The effects of conclusion omission and involvement in persuasion. Journal o f Consumer Research, 15, 225-233. Kardes, F. R. (1993). Consumer inference: Determinants, consequences, and implications for ad­ vertising. In A. A. Mitchell (Ed.), Advertising exposure, memory, and choice (pp. 163-191). Hills­ dale, NJ: Lawrence Erlbaum Associates. Kardes, F. R., & Allen, C. T. (1991). Perceived variability and inferences about brand extensions. Advances in Consumer Research, 18, 392-398. Kardes, F. R., Allen, C. T ., & Pontes, M. J. (1993). Effects of multiple measurement operations on consumer judgment: Measurement reliability or reactivity? Advances in Consumer Research, 20, 280-283. Kardes, F. R., & Gurumurthy, K. (1992). Order-of-entry effects on consumer memory and judg­ ment: An information integration perspective. Journal o f Marketing Research, 29, 343-357. Kardes, F. R., Gurumurthy, K., Chandrashekaran, M ., & Domoff, R. J. (1993). Brand retrieval, consideration set composition, consumer choice, and the pioneering advantage. Journal o f Con­ sumer Research, 20, 62-75. Kardes, F. R., & Herr, P. M. (1990). Order effects in consumer judgment, choice, and memory: The role of initial processing goals. Advances in Consumer Research, 17, 541-546. Kardes, F. R., Kim, J., & Lim, J. (in press). Moderating effects of prior knowledge on the per­ ceived diagnosticity of beliefs derived from implicit versus explicit product claims. Journal o f Business Research.

Kardes, F. R., & Sanbonmatsu, D. M. (1993). Direction of comparison, expected feature correla­ tion, and the set-size effect in preference judgment. Journal o f Consumer Psychology, 2, 39-54. Katz, D. (1960). The functional approach to the study of attitudes. Public Opinion Quarterly, 24, 163-204. Keenan, J. M ., Baillet, S. D., & Brown, P. (1984). The effects of causal cohesion on comprehension and memory. Journal o f Verbal Learning and Verbal Behavior, 23, 115-126. Keller, K. L. (1987). Memory in advertising: The effect of advertising memory cues on brand evalu­ ations. Journal o f Consumer Research, 14, 316-333. Keller, K. L. (1991). Memory and evaluation effects in competitive advertising environments. Jour­ nal o f Consumer Research, 16, 436-476. Kent, R. J., & Allen, C. T. (1993). Competitive interference effects in consumer memory fo r adver­ tising: The role o f brand fam iliarity. Unpublished manuscript. Kiel, G. C., & Layton, R. A. (1981). Dimensions of consumer information seeking. Journal o f Con­ sumer Research, 8, 233-239. Kisielius, J., & Sternthal, B. (1984). Detecting and explaining vividness effects in attitudinal judg­ ments. Journal o f Marketing Research, 21, 54-64. Kisielius, J., & Stemthal, B. (1986). Examining the vividness controversy: An availability-valence interpretation. Journal o f Consumer Research, 12, 418-431. Klar, Y. (1990). Linking structures and sensitivity to judgment-relevant information in statistical and logical reasoning tasks. Journal o f Personality and Social Psychology, 59, 841-858. Klayman, J., & Ha, Y. (1987). Confirmation, disconfirmation, and information in hypothesis test­ ing. Psychological Review, 94, 211-228. Klinger, E. (1975). Consequences of commitment to and disengagement from incentives. Psycho­ logical Review, 82, 1-25.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

459

Koriat, A., Lichtenstein, S., & Fischhoff, B. (1980). Reasons for confidence. Journal o f Experimen­ tal Psychology: Human Learning and Memory, 6 , 107-118. Kraft, R. N. (1987). The influence of camera angle on comprehension and retention of pictorial events. M emory and Cognition, 15, 291-307. Kruglanski, A. W. (1989). Lay epistemics and human knowledge: Cognitive and motivational bases. New York: Plenum. Krugman, H. E. (1965). The impact of television advertising: Learning without involvement. Public Opinion Quarterly, 29, 349-356. Lakatos, I. (1974). Falsification and the methodology of scientific research programs. In I. Lakatos & A. Musgrave (Eds.), Criticism and the growth o f knowledge (pp. 91-195). Cambridge, UK: Cambridge University Press. Larrick, R. P., Morgan, J. N., & Nisbett, R. E. (1990). Teaching the use of cost-benefit reasoning in everyday life. Psychological Science, 1, 362-370. Lauden, L. (1984). Science and values. Berkeley, CA: University of California Press. Levin, I. P., & Gaeth, G. J. (1988). How consumers are affected by the framing of attribute infor­ mation before and after consuming the product. Journal o f Consumer Research, 15, 374-378. Lichtenstein, M ., & Srull, T. K. (1985). Conceptual and methodological issues in examining the relationship between consumer memory and judgment. In L. F. Alwitt & A. A. Mitchell (Eds.), Psychological processes and advertising effects: Theory, research, and application (pp. 113-128). Hillsdale, NJ: Lawrence Erlbaum Associates. Lichtenstein, M ., & Srull, T. K. (1987). Processing objectives as a determinant of the relationship between recall and judgment. Journal o f Experimental Social Psychology, 23, 93-118. Lingle, J. H., & Ostrom, T. M. (1981). Principles of memory and cognition in attitude formation. In R. E. Petty, T. M. Ostrom, & T. C. Brock (Eds.), Cognitive responses in persuasion (pp. 399-420). Hillsdale, NJ: Lawrence Erlbaum Associates. Loken, B., & Ward, J. (1990). Alternative approaches to understanding the determinants of typicali­ ty. Journal o f Consumer Research, 17, 111-126. Lord, C. G., Lepper, M. R., & Preston, E. (1984). Considering the opposite: A corrective strategy for social judgment. Journal o f Personality and Social Psychology, 47, 1231-1243. Lord, C. G., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude polarization: The effects of prior theories on subsequently considered evidence. Journal o f Personality and Social Psychology, 37, 2098-2109. Louviere, J. J. (1988). Analyzing decision making: Metric conjoint analysis. Newbury Park, CA: Sage. Lynch, J. G. (1985). Uniqueness issues in the decompositional modeling of multiattribute over­ all evaluations: An information integration perspective. Journal o f Marketing Research, 22, 1-19. Lynch, J. G., Marmorstein, H., & Weigold, M. F. (1988). Choices from sets including remembered brands: Use of recalled attributes and prior overall evaluations. Journal o f Consumer Research, 15, 169-184. Lynch, J. G., & Srull, T. K. (1982). Memory and attentional factors in consumer choice: Concepts and research methods. Journal o f Consumer Research, 9, 18-37. MacKenzie, S. B. (1986). The role of attention in mediating the effect of advertising on attribute importance. Journal o f Consumer Research, 13, 174-195. MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a media­ tor of advertising effectiveness: A test of competing explanations. Journal o f Marketing Research, 23, 130-143. Madden, T. J., Allen, C. T ., & Twibble, J. L. (1988). Attitude toward the ad: An assessment of diverse measurement indices under different processing “sets.” Journal o f Marketing Research, 25, 242-252. Madigan, S. A. (1969). Intraserial repetition and coding processes in free recall. Journal o f Verbal Learning and Verbal Behavior, 8, 828-835.

46 0

KARDES

Maheswaran, D., Mackie, D. M ., & Chaiken, S. (1992). Brand name as a heuristic cue: The effects of task importance and expectancy confirmation on consumer judgments. Journal o f Consumer Psychology, 7, 317-336. Mandler, G. (1982). The structure of value: Accounting for taste. In M. S. Clark & S. T. Fiske (Eds.), Affect and cognition: The 17th Annual Carnegie Symposium on cognition (pp. 3-36). Hillsdale, NJ: Lawrence Erlbaum Associates. Mandler, G., Nakamura, Y., & Shebo Van Zandt, B. J. (1987). Nonspecific effects of exposure on stimuli that cannot be recognized. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 13, 646-648. Mazursky, D., & Schul, Y. (1988). The effects of advertisement encoding on the failure to discount information: Implications for the sleeper effect. Journal o f Consumer Research, 15, 24-36. McGill, A. L., & Anand, P. (1989). The effect of vivid attributes on the evaluation of alternatives: The role of differential attention and cognitive elaboration. Journal o f Consumer Research, 16, 188-196. McNamara, T. P. (1992). Priming and constraints it places on theories of memory and retrieval. Psychological Review, 99, 650-662. McSweeney, F. K., & Bierley, C. (1984). Recent developments in classical conditioning. Journal o f Consumer Research, 11, 619-631. Meyer, R. J. (1981). A model of multiattribute judgments under attribute uncertainty and informa­ tion constraints. Journal o f Marketing Research, 18, 428-441. Meyer, R. J. (1987). The learning of multiattribute judgment policies. Journal o f Consumer Research, 14, 155-173. Meyers-Levy, J., & Maheswaran, D. (1992). When timing matters: The influence of temporal dis­ tance on consumers’ affective and persuasive responses. Journal o f Consumer Research, 1 9 , 424-433. Meyers-Levy, J., & Peracchio, L. A. (1992). Getting an angle in advertising: The effect of camera angle on product evaluations. Journal o f Marketing Research, 29, 454-461. Meyers-Levy, J., & Tybout, A. M. (1989). Schema congruity as a basis for product evaluation. Journal o f Consumer Research, 16, 39-54. Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81-97. Miniard, P. W ., Bhatla, S., & Rose, R. L. (1990). On the formation and relationship of ad and brand attitudes: An experimental and causal analysis. Journal o f Marketing Research, 27, 290-303. Miniard, P. W ., Sirdeshmukh, D., & Innis, D. E. (1992). Peripheral persuasion and brand choice. Journal o f Consumer Research, 19, 226-239. Mitchell, A. A. (1986). The effects of verbal and visual components of advertisements on brand atti­ tudes and attitude toward the advertisement. Journal o f Consumer Research, 13, 12-24. Mitchell, A. A., & Olson, J. C. (1981). Are product attribute beliefs the only mediator of advertis­ ing effects on brand attitudes? Journal o f Marketing Research, 18, 318-322. Mizerski, R. W. (1982). An attribution explanation of the disproportionate influence of unfavorable information. Journal o f Consumer Research, 9, 301-310. Modigliani, V. (1976). Effects on a later recall by delaying initial recall. Journal o f Experimental Psychology: Human Learning and Memory, 2, 609-622. Moore, D. L., Hausknecht, D ., & Thamodaran, K. (1986). Time compression, response opportuni­ ty, and persuasion. Journal o f Consumer Research, 13, 85-99. Moore, W. L ., & Lehmann, D. R. (1980). Individual differences in search behavior for a nondura­ ble. Journal o f Consumer Research, 7, 296-307. Moser, D. V. (1992). Does memory affect judgment? Self-generated versus recall memory meas­ ures. Journal o f Personality and Social Psychology, 62, 555-563. Nedungadi, P. (1990). Recall and consumer consideration sets: Influencing choice without altering brand evaluations. Journal o f Consumer Research, 17, 263-276. Nedungadi, P., & Hutchinson, J. W. (1985). The prototypicality of brands: Relationships with brand awareness, preference, and usage. Advances in Consumer Research, 12, 498-503.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

461

Nelson, D. L., Schreiber, T. A., & McEvoy, C. L. (1992). Processing implicit and explicit represen­ tations. Psychological Review, 99, 322-348. Nelson, T. O. (1977). Repetition and depth of processing. Journal o f Verbal Learning and Verbal Behavior, 16, 151-172. Newell, A ., & Simon, H. A. (1972). Human problem solving . Englewood Cliffs, NJ: Prentice-Hall. Newman, J., Wolff, W. T., & Hearst, E. (1980). The feature-positive effect in adult human sub­ jects. Journal o f Experimental Psychology: Human Learning and Memory, 6, 630-650. Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings o f social ju d g ­ ment. Englewood Cliffs, NJ: Prentice-Hall. Nisbett, R. E., Zukier, H ., & Lemley, R. E. (1981). The dilution effect: Nondiagnostic information weakens the implications of diagnostic information. Cognitive Psychology, 13, 248-277. Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction deci­ sions. Journal o f Marketing Research, 17, 460-469. Oliver, R. L., & DeSarbo, W. S. (1988). Response determinants in satisfaction judgments. Journal o f Consumer Research, 14, 495-507. Olney, T. J., Holbrook, M. B., & Batra, R. (1991). Consumer responses to advertising: The effects of ad contents, emotions, and attitude toward the ad on viewing time. Journal o f Consumer Research, 17, 440-453. Olshavsky, R. W ., & Granbois, D. H. (1979). Consumer decision making—Fact or fiction? Journal o f Consumer Research, 6, 93-100. Park, C. W ., Iyer, E. S., & Smith, D. C. (1989). The effects of situational factors on in-store grocery shopping behavior: The role of store environment and time available for shopping. Journal o f Con­ sumer Research, 15, 422-433. Park, C. W ., Milberg, S., & Lawson, R. (1991). Evaluation of brand extensions: The role of product feature similarity and brand concept consistency. Journal o f Consumer Research, 18, 185-193. Park, C. W ., & Young, S. M. (1986). Consumer response to television commercials: The impact of involvement and background music on brand attitude formation. Journal o f Marketing Research, 23, 11-24. Pavelchak, M. A., Antil, J. H ., & Munch, J. M. (1988). The Super Bowl: An investigation into the relationship among program context, emotional experience, and ad recall. Journal o f Consumer Research, 15, 360-367. Payne, J. W. (1976). Task complexity and contingent processing in decision making: An informa­ tion search and protocol analysis. Organizational Behavior and Human Decision Processes, 16, 366-387. Payne, J. W ., Bettman, J. R., & Johnson, E. J. (1992). Behavioral decision research: A constructive processing perspective. Annual Review o f Psychology, 43, 87-131. Peterson, D. K., & Pitz, G. F. (1988). Confidence, uncertainty, and the use of information. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 14, 85-92. Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L. Ber­ kowitz (Ed.), Advances in experimental social psychology (Vol. 19, pp. 123-205). New York: Academic Press. Petty, R. E., & Cacioppo, J. T. (1990). Involvement and persuasion: Tradition versus integration. Psychological Bulletin, 107, 367-374. Petty, R. E., Cacioppo, J. T ., & Schumann, D. (1983). Central and peripheral routes to advertis­ ing effectiveness: The moderating role of involvement. Journal o f Consumer Research, 10, 135-146. Petty, R. E., Ostrom, T. M ., & Brock, T. C. (Eds.). (1981). Cognitive responses in persuasion. Hillsdale, NJ: Lawrence Erlbaum Associates. Petty, R. E., Unnava, R. H., & Strathman, A. J. (1991). Theories of attitude change. In T. S. Robertson & H. H. Kassarjian (Eds.), Handbook o f consumer behavior (pp. 241-280). Englewood Cliffs, NJ: Prentice-Hall.

462

KARDES

Petty, R. E., Wells, G. L ., & Brock, T. C. (1976). Distraction can enhance or reduce yielding to propaganda: Thought disruption versus effort justification. Journal o f Personality and Social Psy­ chology, 34, 874-884. Pratkanis, A. R., & Greenwald, A. G. (1989). A sociocognitive model of attitude structure and func­ tion. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22, pp. 245-285). New York: Academic Press. Punj, G. N., & Staelin, R. (1983). A model of consumer search behavior for new automobiles. Jour­ nal o f Consumer Research, 9, 366-380. Raaijmakers, J. G. W ., & Shiffrin, R. M. (1981). Search of associative memory. Psychological Review, 88, 93-134. Ratcliff, R. (1981). Does activation really spread? Psychological Review, 88, 454-462. Ratcliff, R., & McKoon, G. (1988). A retrieval theory of priming in memory. Psychological Review, 95, 385-408. Ratneshwar, S., & Chaiken, S. (1991). Comprehension’s role in persuasion: The case of its moderat­ ing effect on the persuasive impact of source cues. Journal o f Consumer Research, 18, 52-62. Ratneshwar, S., & Shocker, A. D. (1991). Substitution in use and the role of usage context in product category structures. Journal o f Marketing Research, 28, 281-295. Reyes, R. M ., Thompson, W. C ., & Bower, G. H. (1980). Judgmental biases resulting from differ­ ing availabilities of arguments. Journal o f Personality and Social Psychology, 39, 2-12. Ritov, I., & Baron, J. (1992). Status-quo and omission biases. Journal o f Risk and Uncertainty, 5, 49-61. Robinson, W. T., & Fomell, C. (1985). Sources of market pioneering advantages in consumer goods industries. Journal o f Marketing Research, 22, 305-318. Roediger, H. L. (1990). Implicit memory: Retention without remembering. American Psychologist, 45, 1043-1056. Rosch, E., & Mervis, C. B. (1975). Family resemblances: Studies in the internal structure of categories. Cognitive Psychology, 7, 573-605. Ross, W. T ., & Creyer, E. H. (1992). Making inferences about missing information: The effects of existing information. Journal o f Consumer Research, 19, 14-25. Rothbart, M ., Evans, M ., & Fulero, S. (1979). Recall for confirming events: Memory processes and the maintenance of social stereotypes. Journal o f Experimental Social Psychology, 15, 343-355. Rundus, D. (1971). Analysis of rehearsal processes in free recall. Journal o f Experimental Psycholo­ gy, 89, 63-77. Rundus, D. (1973). Negative effects of using list items as recall cues. Journal o f Verbal Learning and Verbal Behavior, 12, 43-50. Russo, J. E. (1977). The value of unit price information. Journal o f Marketing Research, 14, 193-201. Russo, J. E., Staelin, R., Nolan, C. A., Russell, G. J., & Metcalf, B. L. (1986). Nutrition informa­ tion in the supermarket. Journal o f Consumer Research, 13, 48-70. Samuelson, W ., & Zechauser, R. (1988). Status-quo bias in decision making. Journal o f Risk and Uncertainty, 1, 7-59. Sanbonmatsu, D. M ., & Fazio, R. H. (1990). The role of attitudes in memory-based decision mak­ ing. Journal o f Personality and Social Psychology, 59, 614-622. Sanbonmatsu, D. M ., & Kardes, F. R. (1988). The effects of physiological arousal on information processing and persuasion. Journal o f Consumer Research, 15, 379-385. Sanbonmatsu, D. M ., Kardes, F. R., & Gibson, B. D. (1991). The role of attribute knowledge and overall evaluations in comparative judgment. Organizational Behavior and Human Decision Process­ es, 48, 131-146. Sanbonmatsu, D. M ., Kardes, F. R., & Herr, P. M. (1992). The role of prior knowledge and miss­ ing information in multiattribute evaluation. Organizational Behavior and Human Decision Processes, 51, 76-91. Sanbonmatsu, D. M ., Kardes, F. R ., & Sansone, C. (1991). Remembering less and inferring more: The effects of the timing of judgment on inferences about unknown attributes. Journal o f Person­ ality and Social Psychology, 61, 546-554.

8.

CONSUMER JUDGM ENT AND DECISION PROCESSES

463

Sawyer, A. G. (1988). Can there be effective advertising without explicit conclusions? Decide for yourself. In S. Hecker & D. W. Stewart (Eds.), Nonverbal communications in advertising (pp. 159-184). Lexington, MA: Lexington. Sawyer, A. G., & Howard, D. J. (1991). Effects of omitting conclusions in advertisements to in­ volved and uninvolved audiences. Journal o f Marketing Research, 28, 467-474. Schmalensee, R. (1982). Product differentiation advantages to pioneering brands. American Economic Review, 72, 159-180. Schul, Y., & Mazursky, D. (1990). Conditions facilitating successful discounting in consumer deci­ sion making. Journal o f Consumer Research, 16, 442-451. Schumann, D. W ., Petty, R. E., & Clemons, D. S. (1990). Predicting the effectiveness of different strategies of advertising variation: A test of the repetition-variation hypotheses. Journal o f Con­ sumer Research, 17, 192-202. Schwartz, S. H. (1978). Temporal instability as a moderator of the attitude-behavior relationship. Journal o f Personality and Social Psychology, 36, 715-724. Shaklee, H ., & Fischhoff, B. (1982). Strategies of information search in causal analysis. Memory and Cognition, 14, 150-163. Shedler, J., & Manis, M. (1986). Can the availability heuristic explain vividness effects? Journal o f Personality and Social Psychology, 51, 26-36. Sheppard, B. H., Hartwick, J., & Warshaw, P. R. (1988). The theory of reasoned action: A meta­ analysis of past research with recommendations for modifications and future research. Journal o f Consumer Research, 15, 325-343. Sherman, S. J., & Corty, E. (1984). Cognitive heuristics. In R. S. Wyer & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 2, pp. 189-286). Hillsdale, NJ: Lawrence Erlbaum Associates. Sherman, S. J., McMullen, M. N., & Gavanski, I. (1992). Natural sample spaces and the inversion of conditional judgments. Journal o f Experimental Social Psychology, 28, 401-421. Sherman, S. J., Skov, R. B., Hervitz, E. F., & Stock, C. B. (1981). The effects of explaining hypothet­ ical future events: From possibility to probability to actuality and beyond. Journal o f Experimental Social Psychology, 17, 142-158. Sherman, S. J., Zehner, K. S., Johnson, J., & Hirt, E. R. (1983). Social explanation: The role of timing, set, and recall on subjective likelihood estimates. Journal o f Personality and Social P sy­ chology, 44, 1127-1143. Shiffrin, R. M. (1988). Attention. In R. C. Atkinson, R. J. Herrnstein, G. Lindzey, & R. D. Luce (Eds.), Stevens’ handbook o f experimental psychology (pp. 739-811). New York: Wiley. Shimp, T. A., Stuart, E. W ., & Engle, R. W. (1991). A program of classical conditioning experi­ ments testing variations in the conditioned stimulus and contents. Journal o f Consumer Research, 18, 1-12. Shugan, S. M. (1980). The cost of thinking. Journal o f Consumer Research, 7, 99-111. Simmons, C. J., & Lynch, J. G. (1991). Inference effects without inference making? Effects of missing information on discounting and use of presented information. Journal o f Consumer Research, 17, 477-491. Simonson, I. (1989). Choice based on reasons: The case of attraction and compromise effects. Jour­ nal o f Consumer Research, 16, 158-174. Simonson, I. (1992). The influence of anticipating regret and responsibility on purchase decisions. Journal o f Consumer Research, 19, 105-118. Simonson, I., Huber, J., & Payne, J. (1988). The relationship between prior brand knowledge and information acquisition order. Journal o f Consumer Research, 14, 566-578. Simonson, I., & Tversky, A. (1992). Choice in context: Tradeoff contrast and extremeness aver­ sion. Journal o f Marketing Research, 29, 281-295. Sivacek, J., & Crano, W. D. (1982). Vested interest as a moderator of attitude-behavior consisten­ cy. Journal o f Personality and Social Psychology, 43, 210-221. Skowronski, J. J., & Carlston, D. E. (1987). Social judgment and social memory: The role of cue diagnosticity in negativity, positivity, and extremity biases. Journal o f Personality and Social Psy­ chology, 52, 689-699.

464

KARDES

Skowronski, J. J., & Carlston, D. E. (1989). Negativity and extremity biases in impression forma­ tion: A review of explanations. Psychological Bulletin, 105 , 131-142. Sloan, L. R., & Ostrom, T. M. (1974). Amount of information and interpersonal judgment. Journal o f Personality and Social Psychology, 29, 23-29. Smith, E. E., Langston, C ., & Nisbett, R. E. (1992). The case for rules in reasoning. Cognitive Science, 16, 1-40. Smith, M. B., Bruner, J. S., & White, R. W. (1956). Opinions and personality. New York: Wiley. Smith, R. E., & Swinyard, W. R. (1983). Attitude-behavior consistency: The impact of product trial versus advertising. Journal o f Marketing Research, 20, 257-267. Snyder, M. (1974). Self-monitoring of expressive behavior. Journal o f Personality and Social Psy­ chology, 30, 526-537. Snyder, M ., & Swann, W. B. (1978). Hypothesis-testing in social interaction. Journal o f Personality and Social Psychology, 36, 1202-1212. Srinivasan, N., & Ratchford, B. T. (1991). An empirical test of a model of external search for au­ tomobiles. Journal o f Consumer Research, 18, 233-242. Srull, T. K. (1982). The representation of consumer information in memory. Advances in Consumer Research, 9, 499-501. Srull, T. K. (1984). Methodological techniques for the study of person memory and social cognition. In R. S. Wyer & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 2, pp. 1-72). Hillsdale, NJ: Lawrence Erlbaum Associates. Srull, T. K., Lichtenstein, M ., & Rothbart, M. (1985). Associative storage and retrieval processes in person memory. Journal o f Experimental Psychology: Learning, Memory, and Cognition, 11, 316-345. Srull, T. K., & Wyer, R. S. (1979). The role of category accessibility in the interpretation of infor­ mation about persons: Some determinants and implications. Journal o f Personality and Social Psy­ chology, 37, 1660-1672. Srull, T. K., & Wyer, R. S. (1980). Category accessibility and social perception: Some implications for the study of person memory and interpersonal judgments. Journal o f Personality and Social Psychology, 38, 841-856. Stayman, D. M ., & Aaker, D. A. (1988). Are all effects of ad-induced feelings mediated by Aad? Journal o f Consumer Research, 15, 368-373. Stayman, D. M ., Alden, D. L ., & Smith, K. H. (1992). Some effects of schematic processing on consumer expectations and disconfirmation judgments. Journal o f Consumer Research, 19, 240-255. Stayman, D. M ., & Kardes, F. R. (1992). Spontaneous inference processes in advertising: Effects of need for cognition and self-monitoring on inference generation and utilization. Journal o f Con­ sumer Psychology, 1, 125-142. Stuart, E. W ., Shimp, T. A., & Engle, R. W. (1987). Classical conditioning of consumer attitudes: Four experiments in an advertising context. Journal o f Consumer Research, 14, 334-349. Sujan, M. (1985). Consumer knowledge: Effects on evaluation strategies mediating consumer judg­ ments. Journal o f Consumer Research, 12, 31-46. Sujan, M ., & Dekleva, C. (1987). Product categorization and inference making: Some implications for comparative advertising. Journal o f Consumer Research, 14, 372-378. Taylor, S. E., & Thompson, S. C. (1982). Stalking the elusive “vividness” effect. Psychological Review, 89, 155-181. Thaler, R. (1980). Toward a positive theory of consumer choice. Journal o f Economic Behavior and Organization, 1, 39-60. Trope, Y. (1986). Identification and inferential processes in dispositional attribution. Psychological Review, 93, 239-257. Troutman, C. M ., & Shanteau, J. (1976). Do consumers evaluate products by adding or averaging attribute information? Journal o f Consumer Research, 3, 101-106. Tulving, E ., & Pearlstone, Z. (1966). Availability versus accessibility of information in memory for words. Journal o f Verbal Learning and Verbal Behavior, 5, 381-391.

8.

CONSUMER JUDGMENT AND DECISION PROCESSES

46 5

Tversky, A. (1977). Features of similarity. Psychological Review, 84, 327-353. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 1124-1131. Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211, 453-458. Tversky, A., & Kahneman, D. (1991). Loss aversion in riskless choice: A reference-dependent model. Quarterly Journal o f Economics, 106, 1040-1061. Tversky, A., Sattah, S., & Slovic, P. (1988). Contingent weighting in judgment and choice. Psycho­ logical Review, 95, 371-384. Tybout, A. M ., Calder, B. J., & Sternthal, B. (1981). Using information processing theory to design marketing strategies. Journal o f Marketing Research, 18, 73-79. Tybout, A. M ., & Scott, C.; A. (1983). Availability of well-defined internal knowledge and the atti­ tude formation process: Information aggregation versus self-perception. Journal o f Personality and Social Psychology, 44, 474-491. Tybout, A. M ., Sternthal, B., & Calder, B. J. (1983). Information availability as a determinant of multiple request effectiveness. Journal o f Marketing Research, 20, 280-290. Tyler, S. W ., Hertel, P. T ., McCallum, M. C., & Ellis, H. C. (1979). Cognitive effort and memory. Journal o f Experimental Psychology: Human Learning and Memory, 5, 607-617. Unnava, R. H., & Bumkrant, R. E. (1991). An imagery-processing view of the role of pictures in print advertisements. Journal o f Marketing Research, 28, 226-231. Urban, G. L., Carter, T ., Gaskin, S., & Mucha, Z. (1986). Market share rewards to pioneering brands: An empirical analysis and strategic implications. Management Science, 32, 645-659. Urbany, J. E. (1986). An experimental examination of the economics of information. Journal o f Con­ sumer Research, 13, 257-271. Urbany, J. E ., Dickson, P. R., & Wilkie, W. L. (1989). Buyer uncertainty and information search. Journal o f Consumer Research, 16, 208-215. Venkatraman, M. P., Marlino, D., Kardes, F. R., & Sklar, K. B. (1990). The interactive effects of message appeal and individual differences on information processing and persuasion. Psycholo­ gy and Marketing, 7, 85-96. von Neumann, J., & Morgenstem, O. (1944). Theory o f games and economic behavior. Princeton, NJ: Princeton University Press. Walker, N. (1986). Direct retrieval from elaborated memory traces. Memory and Cognition, 14, 321-328. Wason, P. C. (1960). On the failure to eliminate hypotheses in a conceptual task. Quarterly Journal o f Experimental Psychology, 12, 129-140. Weiner, B. (1985). “Spontaneous” causal thinking. Psychological Bulletin, 97, 74-84. Wilkie, W. L., & Pessemier, E. (1973). Issues in marketing’s use of multi-attribute attitude models. Journal o f Marketing Research, 10, 428-441. Woodruff, R. B., Cadotte, E. R., & Jenkins, R. (1983). Modeling consumer satisfaction processes using experience-based norms. Journal o f Marketing Research, 20, 296-304. Wright, P. (1974). The harassed decision maker: Time pressures, distractions, and the use of evi­ dence. Journal o f A pplied Psychology, 59, 555-561. Wright, P. (1980). Message-evoked thoughts: Persuasion research using thought verbalizations. Journal o f Consumer Research, 7, 151-175. Wyer, R. S. (1970). Information redundancy, inconsistency, and novelty and their role in impres­ sion formation. Journal o f Experimental Social Psychology, 6, 111-127. Wyer, R. S. (1974). Cognitive organization and change: An information processing approach. Hills­ dale, NJ: Lawrence Erlbaum Associates. Wyer, R. S. (1977). The role of logical and nonlogical factors in making inferences about category membership. Journal o f Experimental Social Psychology, 13, 577-595. Wyer, R. S., & Carlston, D. E. (1979). Social cognition, inference, and attribution. Hillsdale, NJ: Lawrence Erlbaum Associates.

466

KARDES

Wyer, R. S., & Gordon, S. E. (1984). The cognitive representation of social information. In R. S. Wyer & T. K. Srull (Eds.), Handbook o f social cognition (Vol. 2, pp. 73-150). Hillsdale, NJ: Lawrence Erlbaum Associates. Wyer, R. S., & Srull, T. K. (1986). Human cognition in its social context. Psychological Review, 93, 322-359. Wyer, R. S., & Srull, T. K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Lawrence Erlbaum Associates. Yalch, R. F ., & Elmore-Yalch, R. (1984). The effect of numbers on the route to persuasion. Journal o f Consumer Research, 11, 522-527. Yi, Y. (1990). The effects of contextual priming in print advertisements. Journal o f Consumer Research, 17, 215-222. Zaichowsky, J. (1985). Measuring the involvement construct. Journal o f Consumer Research, 12, 341-352.

Author Index

Note: This index refers the reader to pages in both volumes of this Handbook. Volume numbers (1), (2) are given in parentheses. Page numbers in italics indicate the location of full bibliographical

references.

A Aaker, D. A., (2) 407, 431, 451, 464 Abbott, R. A., (2) 264, 286 Abelson, R. P., (1) 42, 44, 76, 77, 89, 91, 96, 132, 133, 144, 149, 187, 190, 192, 194, 205, 332, 363, 404, 413. (2) 71, 73, 81, 120, 130, 176, 179, 205, 2 3 7 Abraham, L. M ., (2) 342, 386 Abramson, L. Y., (1) 357, 398, 404, 405, 414. (2) 293, 294, 299, 301, 307, 311, 317, 320, 326, 327, 330, 334 Abu-Zahra, H ., (2) 244, 284 Achee, J. W ., (1) 390, 391, 397, 399, 402, 411

Ackerman, P. L., (2) 341, 342, 391 Ackerman, R., (2) 299, 311, 3 26 Acklin, M ., (2) 256, 282 Acorn, D. A., (1) 223, 283 Adams, H. E ., (1) 380, 406 Adams, S., (2) 374, 394 Adams, S. C., (2) 400, 453 Adelman, P. K., (1) 385, 404 Adler, A., (1) 193, 194 Adler, J. F., (1) 308, 31 7 Adler, R. F., (1) 188, 208 Adorno, T. E., (2) 145, 179 Affleck, G., (2) 270, 271, 279 Agnoli, F., (1) 308, 3 1 7 Agre, P. E ., (1) 133, 144 Agustinelli, G., (2) 428, 451

Agustsdottir, S., (1) 127, 149 Ahadi, S., (1) 361, 404 Ahlering, R. F., (2) 92, 120 Ahlm, K., (1) 312, 320 Ahrens, A. H ., (2) 307, 3 2 7 Aiken, M ., (2) 70, 132 Ainsworth, M. D. S., (2) 191, 193, 231 Aird, P., (2) 51, 52, 61 Ajzen, I., (2) 77-79, 110, 111, 116, 117, 120, 126, 164, 172, 173, 181, 449, 451, 455

Alba, J. W ., (1) 86, 92. (2) 83, 84, 93, 106, 120, 123, 376, 379, 386, 401, 402, 408, 409, 412, 415-417, 420, 422, 423, 430, 433, 451, 453 Albert, R. D., (2) 87, 127 Albert, S., (2) 271, 279 Albright, J. S., (1) 157, 200 Albright, L ., (2) 309, 333 Alden, D. L., (2) 406, 464 Aldrich, J. H ., (2) 170, 185 Aldwin, C., (2) 318, 329 Alexander, R. A., (2) 346, 354, 358, 370, 379, 380, 390, 393, 394 Alfieri, T ., (1) 27, 39 Allan, L. G., (1) 140, 142, 147 Allen, C. T ., (2) 412, 420, 422, 431, 451, 458, 459

Allen, R. B., (1) 90, 92 Allen, S. W ., (1) 140, 144 Allison, S. T ., (2) 56, 58, 63 467

468

AUTHOR INDEX

Alloy, L. B., (1) 357, 398, 404, 405. (2) 290, 292-299, 306, 307, 311-317, 320, 326, 327, 329, 331, 332, 334, 337, 338

Allport, G. W ., (1) 159, 160, 194, 194. (2) 4, 8, 42, 53, 58 Allred, K. D., (2) 264, 266, 280, 286 Altom, M. W ., (1) 140, 148, 160, 163, 202, 243, 280. (2) 23, 64 Alumbaugh, M. J., (2) 244, 248, 285 Alva, A. L ., (2) 257, 280 Anand, P., (2) 407, 460 Andersen, S. M ., (1) 8, 26, 31, 32, 13, 92, 101, 144, 234, 246, 258, 275, 279, 282. (2) 20, 35, 58, 307, 308, 311, 314, 325, 3 2 7 Anderson, A., (1) 291, 3 1 7 Anderson, B. L ., (2) 251, 252, 280 Anderson, C. A., (1) 86, 87, 92, 97, 215, 249, 275. (2) 37, 66, 87, 120, 291, 327, 377, 385, 395 Anderson, J. R., (1) 49, 55, 86, 92, 99-101, 103, 105-108, 110, 112,113, 116-119, 122, 129, 133, 134, 138, 144, 145, 148, 150, 179-181, 186, 190, 194, 195, 368, 372, 374, 405. (2) 80, 120, 164, 180, 256, 280, 298, 327, 416, 418, 421, 426, 438, 451 Anderson, K., (1) 396, 415. (2) 163, 184 Anderson, N. H ., (1) 17, 32, 56, 92, 217, 218, 275. (2) 85, 94, 111, 120, 137, 164, 172, 180, 420, 423-426, 428, 429, 451, 452 Anderson, R. C ., (2) 166, 180 Andrews, B., (2) 291, 32 7 Ansolabehere, S., (2) 151-153, 180 Antil, J. H., (2) 402, 412, 461 Antoni, M. H., (2) 265, 280 Antoniou, A. A., (1) 343-348, 365, 413 Anzai, Y., (1) 107, 145 Appelbaum, G. A ., (2) 404, 454 Arkes, H. R., (2) 79, 120, 409, 447, 452 Armenakis, A. A., (2) 370, 38 7 Armour, V., (1) 142, 145 Armstrong, G. M ., (2) 93, 122 Arnold, A., (2) 244, 284 Amoult, L. H ., (2) 291, 3 2 7 Aron, A., (1) 136, 145. (2) 221, 233 Aron, E. N., (1) 136, 145 Aronson, E., (1) 42, 43, 91. (2) 96, 120, 140 Arvey, R. D., (2) 342, 386 Asch, S. E., (1) 217, 275. (2) 102, 120 Ashbrook, P. W ., (1) 370-372, 375, 388, 400, 40 7

Ashby, F. G., (1) 139, 145 Ashmore, R. D ., (2) 2, 15, 18, 58, 62, 367, 389 Ashton, W ., (1) 218, 280 Aspinwall, L. G., (2) 228, 237, 264, 270-272, 287, 291, 3 3 7 Asuncion, A. G., (2) 43, 56, 63, 84, 100, 131 Atkin, R. S., (2) 347, 386, 3 8 7 Atman, C. J., (2) 251, 260, 284 Audrain, P. C., (1) 382, 384, 410 Austin, G. A., (1) 158, 195 Austin, J. T ., (2) 344-346, 386 Austin, W ., (2) 198, 201, 231 Averill, J. R., (1) 326, 405. (2) 292, 3 2 7 Avolio, B. J., (2) 346, 366, 368, 396 Axsom, D., (2) 92, 120

B Baddeley, A. D., (1) 185, 201, 373, 375, 409 Bagley, C., (2) 263, 284 Bagozzi, R. P., (2) 116, 117, 120 Bailey, S., (1) 396, 415 Baker, S. M ., (2) 7, 18, 51, 52, 59, 74, 100, 103, 121, 135, 406, 455 Baker, W ., (2) 415, 452 Baillet, S. D., (1) 176, 200. (2) 313, 332, 408, 458 Baldwin, M. W ., (1) 9, 12, 32. (2) 81, 105, 112, 123

Balzer, W. K., (2) 3 4 5-341, 353, 355, 356, 358, 359, 370, 371, 373, 379, 384, 386, 3 9 2-394, 39 6

Banaji, M. R., (1) 169, 176, 177, 197. (2) 41, 58, 291, 3 2 7 Bandura, A., (2) 290, 3 2 7 Bane, C. M ., (2) 95, 133 Banks, C. A., (2) 342, 386 Barber, J. D ., (2) 145, 180 Bard, P. A., (1) 327, 405 Barden, R. C., (1) 398, 399, 411 Barefoot, J. C ., (2) 266, 288 Barger, S. D., (2) 251, 254, 281 Bargh, J. A., (1) 2-8, 10-17, 19-21, 22, 24-31, 32-36, 38, 39, 44, 53, 54, 58,92, 94, 95, 100, 105, 114, 115, 117, 123, 125, 129, 131, 132, 134, 136, 141, 144, 145, 147, 163, 195, 254, 256, 263, 264, 275, 278, 281, 305, 318, 381, 409. (2) 8, 40, 58, 61, 70, 83, 121, 154, 161, 166, 167, 177, 180, 257, 280, 300, 304, 305, 307, 308, 312, 314, 327, 336, 342, 348, 351, 365, 368, 373, 379, 385, 386, 416, 456

AUTHOR INDEX

Barlow, D. H., (2) 290, 292, 297, 301, 317, 327, 329

Baron, J., (1) 219, 283. (2) 446, 447, 462 Baron, R., (1) 186, 2 0 3 , 218, 280 Baron, R. S., (1) 255, 275, 279. (2) 44, 62, 319, 328 Barrett, K. C., (1) 330, 406 Barry, B., (2) 346, 355, 360, 396 Barry, S. R., (2) 346, 355, 395 Barsalou, L. W ., (1) 77, 92, 102, 145, 163, 177, 191-193, 195. (2) 21, 58, 298, 327, 348, 371, 387, 416, 452 Bartels, L. M ., (2) 152, 153, 158, 180 Bartholomew, K., (2) 194, 195, 231 Bartlett, F. C ., (1) 76, 86, 92 Bartlett, J. C., (I) 375, 405 Bassili, J. N., (1) 24, 33, 51, 92, 267, 275. (2) 81, 115, 121 Bassock, M ., (1) 306, 321 Batra, R., (2) 86, 93, 101, 102, 121, 402, 407, 411, 412, 452, 461 Batson, C. D., (1) 326, 405 Baucom, D. H., (2) 222-224, 231, 233 Baumann, L. J., (2) 249, 253, 275, 280 Baumeister, R. F., (1) 128, 145, 167, 191. (2) 91, 128 Baumgardner, M. H., (2) 113, 114, 135, 136, 167, 183 Baumgartner, A. H ., (1) 154, 195 Bazerman, M ., (2) 340, 347, 377, 386, 394 Beach, L. R., (2) 352, 388 Beach, R., (1) 233, 278 Beach, S. R., (2) 223-225, 231-2 3 4 Beattie, A. E., (1) 71, 98. (2) 39, 60, 19, 140

46 9

Belleza, F. S., (1) 169, 176, 195, 197 Bellows, N., (1) 11, 3 7 Belmore, S. M ., (2) 422, 440, 452 Bern, D. J., (1) 11, 33, 127, 145, 181, 195. (2) 70, 75, 89, 104, 105, 121, 412, 437, 452 Bern, S. L., (2) 365, 38 7 Bemmels, B., (2) 363, 38 7 Bentler, P. M ., (2) 291, 33 7 Berelson, B., (2) 146, 165, 180, 183 Berenstein, V., (2) 255, 286 Berent, M ., (2) 81, 121, 130 Berg, J. H., (2) 197, 219, 231 Berger, I. E., (2) 448, 452 Berger, S. M ., (2) 214, 231 Berglass, S., (2) 105, 129 Berkowitz, L., (1) 154, 195, 237, 238, 275, 330, 405. (2) 97, 121 Berley, R. A., (2) 223, 235 Berman, J. S., (2) 310, 338 Berman, L., (1) 312, 320 Bernard, T. M ., (2) 253, 254, 283 Bernardin, H. J., (2) 378, 383, 385, 387 Bernstein, S., (1) 166, 202 Berntson, G. G., (2) 72, 108, 123 Berry, D. S., (1) 76, 92, 218, 275 Berscheid, E., (1) 218, 229, 240, 276, 283, 363, 405. (2) 33 , 66, 196, 197, 205, 206, 218, 220, 222, 228-230, 231, 235, 237, 238

Bersoff, D. N., (2) 362, 389 Bettman, J. R., (2) 174, 185, 399, 401, 417-419, 424, 445, 452, 461 Betz, A. L., (1) 9, 11, 36. (2) 109, 130 Beyth-Marom, R., (2) 444, 455 Bhatla, S., (2) 91, 100, 133, 407, 460 Biehel, G. J., (2) 417, 418, 427, 430, 452,

Beatty, S. E., (2) 401, 417, 452 Beavin, J. H., (1) 72, 97 Bechtold, A., (2) 24, 39, 58 454 Beck, A. H., (1) 114, 145 Biek, M ., (2) 81, 93, 141 Beck, A. T ., (1) 114,145. (2)290, 292, Bierer, L. M ., (1) 370, 413 293, 297, 298, 300, 301,307,313, Bierly, C., (2) 108, 133, 412, 453, 460 317, 327, 328, 331 Biemat, M ., (2) 33, 39, 58 Becker, B., (2) 347, 353, 355-357, 359, 38 7 Billig, M ., (2) 8, 58, 67 Becker, M ., (2) 244, 292 Billman, D., (2) 361, 38 7 Becker, S. E., (2) 299, 335 Bimbaum, D., (2) 261, 262, 275, 285 Beckwith, N. E., (2) 430, 452 Bimbaum, M ., (2) 111, 121 Beekum, R., (2) 377, 38 7 Birrell, P., (1) 74, 96, 214, 216, 260, 281. Beggs, I., (1) 142, 145 (2) 5, 22, 24, 65 Behr, R. L ., (2) 156, 180 Bishop, G. D., (2) 249, 253-255, 257, 275, Beike, D. R., (1) 43 280 Belch, G. E., (2) 407, 459 Bishop, G. F., (2) 155, 180

470

AUTHOR INDEX

Bissonnette, V ., (1) 76, 95 Black, J. B., (1) 76, 77, 92, 190, 192, 795, 205

Blalock, S., (2) 270, 282 Blanchard, E., (2) 244, 282 Blanchard, I., (2) 270, 282 Blaney, P. H ., (1) 370, 373, 376, 377, 380, 405. (2) 101, 727, 266, 28 0 , 299, 300, 328 Blank, A ., (1) 2, 3 6 Blankstein, K. R., (2) 319, 330 Blascovich, J., (2) 72, 727, 725 Blatt, S. J., (2) 291, 304, 328 Blaxton, T. A ., (1) 121, 122, 149, 184, 205 Blehar, M. L ., (2) 191, 193, 231 Blencoe, A. G., (2) 372, 3 9 7 Bless, H ., (1) 14, 15, 28, 38, 141, 150, 290, 291, 298, 303, 317, 320, 321, 379, 385, 389, 392-395, 398, 405, 414. (2) 97, 100, 101, 121, 137, 139, 262, 281, 343, 363, 371, 373, 376, 385, 395

Bleuler, E., (1) 158, 164, 795 Blevans, S. A., (2) 96, 126 Bloom, H. S., (2) 149, 180 Blovin, D. C., (1) 188, 208 Blumer, C., (2) 447, 452 Bobbitt, S. A., (2) 243, 282 Bodenhausen, G. V ., (1) 42, 57, 60, 76, 81-83, 92, 98, 128, 145, 190, 208, 223, 254-256, 270, 275, 284, 394, 405. (2) 24, 33-37, 39, 40, 42-45, 58, 63, 119, 727, 229, 234, 431, 453 Boehm, L. E., (2) 79, 120, 409, 452 Boettger, R., (1) 254, 284, 401, 402, 416. (2) 95, 139, 373, 376, 39 6 Boggs, S. R., (2) 263, 28 6 Bohner, G., (1) 279, 385, 389, 392-395, 405, 414. (2) 100, 101, 121, 137 Bohnert, M ., (2) 297, 328 Bois, J., (2) 370, 394 Bollenbach, A. K., (1) 376, 411 Bollnow, U. F., (1) 384, 405 Bolton, W ., (1) 292, 334 Boman, M ., (1) 166, 204 Bond, C. F., (2) 291, 332 Bond, R. N., (1) 5, 8, 10, 11, 32, 53 , 92, 169, 799, 360, 409. (2) 8, 58, 166, 180, 297, 331 Boninger, D. S., (2) 81, 113, 121, 130 Boon, C ., (2) 291, 33 6

Booth-Kewley, S., (2) 265, 282 Borden, R. J., (2) 211, 212, 232 Borgida, E., (1) 137, 145, 253, 254, 279. (2) 38, 39, 63, 81, 91, 116, 727, 128, 140, 170, 185, 346, 361, 362, 389, 396, 405, 453, 45 7 Borman, C., (1) 26, 39, 116, 117, 150 Borman, W. C., (2) 346, 352, 358, 363, 366, 367, 379, 380, 382, 387, 394 Bornstein, R. F., (1) 9-11, 33, 141, 145. (2) 110, 727 Borys, R., (1) 12, 38 Bostrom, A., (2) 251, 260, 284 Bottger, P. C., (2) 340, 3 8 7 Bouchard, T. J., (2) 342, 386 Bourne, L. E., (2) 444, 454 Bovair, S., (1) 107, 112, 147 Bower, G. H ., (1) 8, 34, 44, 55, 77, 92, 176, 179-181, 184, 190, 795, 372-376, 379, 380, 405, 406, 408, 409. (2) 7, 21, 58, 94, 101, 727, 174, 175, 180, 216, 218, 231, 234, 256, 262, 280, 313, 328, 418, 435, 451, 462

Bowlby, J., (2) 191, 231, 232 Boyes-Brehm, P., (1) 159, 162, 205 Bozzolo, A., (2) 92, 107, 727 Bradbum, N. M ., (1) 299, 302, 317, 321 Bradbury, T. N., (2) 222, 223, 225, 232, 234, 290, 291, 330 Bradley, B., (1) 376, 411. (2) 291, 328 Bradshaw, D., (2) 193, 238 Brady, H. E., (2) 151, 153, 180 Braly, K., (2) 49, 62 Branwell, V. H ., (2) 244, 284 Brand, S., (1) 334, 342, 406 Brannick, J., (2) 380, 3 8 7 Branscombe, N., (1) 6, 14, 26, 38, 39, 116, 117, 121, 122, 136, 150, 256, 275 Bransford, J. D ., (1) 112, 122, 148, 371, 415 Braver, M ., (2) 81, 129 Braverman, D. L ., (1) 383, 411. (2) 102, 132 Bray, R. M ., (2) 103, 722 Breckler, S. J., (1) 161, 795. (2) 71, 72, 79, 100, 101, 722, 135, 176, 180 Brehm, J. W ., (1) 10, 40. (2) 95, 722, 294, 338

Brehm, S. S., (2) 290, 328 Brehmer, B., (2) 443, 453 Brekke, N., (1) 253, 254, 279. (2) 38, 39, 63 Brengelman, D ., (1) 167, 198

AUTHOR INDEX

Brett, J. F., (2) 265, 280 Brewer, M. B., (1) 21, 23, 24, 28, 33, 43, 73, 92, 125, 130, 131, 138, 145, 170, 195, 216, 219, 226, 250-253, 257, 269, 271, 276, 280, 281. (2) 4, 6, 7, 9, 17, 18, 21, 25, 27, 29, 31, 32, 39, 40, 42, 50-52, 55, 58, 59, 62, 64, 65, 302, 328, 344, 387, 429, 431, 453 Brewer, W. F., (1) 47, 84, 92, 156, 179, 195. (2) 246, 280, 293, 32S, 348, 557 Brewin, C. R., (2) 290, 291, 306, 307, 327, 328, 330

Brice, P., (I) 87, 128 Brickman, P., (2) 227, 232 Briede, C., (2) 240, 253, 275, 280 Brief, A. P., (2) 265, 280 Brigham, J. C., (2) 15, 59 Brislin, R. W ., (1) 174, 20 7 Britt, S. H., (2) 400, 453 Broadbent, D. E., (1) 392, 406 Brock, T. C., (1) 254, 281. (2) 81, 83, 85, 86, 90, 93, 94, 107, 113, 121, 122, 134, 135, 137, 156, 185, 402, 411, 414, 423, 461, 462 Brody, N., (1) 9, 38 Brody, R. A., (2) 147, 150, 180, 186 Brooks, L. R., (1) 140, 144. (2) 20, 59 Brooks-Gunn, J., (1) 172, 202 Broughton, J. M ., (1) 172, 195 Brown, A. S., (1) 119, 141, 148 Brown, D., (2) 243, 245, 284 Brown, G., (2) 297, 327, 328 Brown, G. A., (2) 292, 328 Brown, J., (2) 41, 62 Brown, J. D., (1) 128, 151. (2) 291, 3 3 7 Brown, N. R., (1) 191, 195 Brown, P., (1) 309-311, 317, 408, 458 Brown, R., (1) 178, 195. (2) 31, 65 Brown, S. C ., (2) 163, 185 Brown, S. P., (2) 401, 407, 453, 457 Brucks, M ., (2) 93, 122, 401, 402, 412, 453 Bruner, J. S., (1) 158, 195, 392, 406. (2) 7, 21, 59, 72, 138, 166, 180, 402, 404, 413, 414, 453, 464 Bryant, J., (1) 11, 40 Buck, R., (2) 228, 232 Buckley, M. R., (2) 383, 3 8 7 Buckley, R. M ., (2) 370, 3 8 7 Budesheim, T. L., (1) 58, 65-69, 90, 98. (2) 172, 179, 187, 372, 3 9 7 Bugental, J. F. T ., (1) 168, 196

471

Bundy, R. P., (2) 8, 67 Burbach, D. J., (2) 251, 280 Burgess, M. L., (1) 255, 275. (2) 319, 328 Burgoon, M ., (2) 115, 135 Burke, J. J., (2) 265, 280 Burke, P. A., (1) 162, 196 Burke, P. T., 91) 162, 196 Burke, R. R., (2) 93, 122, 418-420, 453 Burkrant, R. E., (2) 90, 91, 122, 405, 465 Burleson, G., (1) 375, 405 Bumstein, E., (2) 84, 97, 98, 122, 137 Burton, H ., (1) 176, 177, 184, 201 Burton, T ., (2) 94, 140 Bush, H ., (2) 244, 284 Bush, L. K., (1) 331, 406 Butler, G., (2) 301, 314, 328 Butterfield, E. C., (2) 160, 183 Buttram, R. T ., (1) 26, 39, 116, 117, 120, 130, 136, 143, 150, 220, 283 Buunk, B. P., (2) 202-204, 236, 270, 272, 280, 287, 291, 33 7 Byrne, D., (1) 386, 406

c Cacioppo, J. T., (1) 29, 33, 44, 54, 96, 124, 149, 265, 281, 331, 393, 413. (2) 70-73, 81, 84-86, 88, 90-96, 98-100, 104, 106-108, 110-113, 116, 117, 119, 122, 123, 127, 131, 134-136, 111, 185, 251, 252, 280, 344, 376, 382, 394, 402, 411-413, 434, 453, 461

Cadotte, R. R., (2) 406, 465 Cafferty, T. P., (2) 347, 350, 351, 372, 374, 375, 385, 387, 388, 396, 3 9 7 Cafry, D., (2) 380, 3 9 7 Calder, B. J., (2) 415, 418, 437, 438, 453, 464

Cameron, L. D., (2) 249, 253, 275, 280 Camp, C., (1) 371, 413 Campbell, A., (1) 289, 317. (2) 148, 165, 180

Campbell, B., (2) 116, 121 Campbell, D. T ., (2) 8, 59, 366, 392 Campbell, J., (2) 213, 238, 271, 28 7 Campbell, J. D., (2) 306, 309, 328 Campbell, J. P., (2) 367, 394 Camper, P. M ., (2) 223, 232 Campos, J. J., (1) 330, 406

472

AUTHOR INDEX

Cann, A., (2) 437, 453 Cantor, N., (1) 11, 21, 57, 132, 134, 137, 138, 150, 156, 157, 160, 162, 163, 196, 200, 204. (2) 17, 59, 293, 329, 343, 395 Cantril, H., (1) 297, 319 Cantu, T ., (2) 257, 280 Capon, N., (2) 424, 452 Caputo, C ., (1) 304, 319 Cardone, S. S., (1) 188, 20 7 Cardy, R. L., (2) 347, 353, 355-357, 359, 364, 377-379, 387, 388 Carlisle, C. J., (2) 44, 63 Carlsmith, J. M ., (1) 10, 34. (2) 96, 126 Carlson, R. A ., (1) 113, 119, 145 Carlston, D. E ., (1) 42, 44-46, 51-54, 56, 59, 60, 65, 70, 73, 76, 88-91, 93, 94, 97, 98, 103, 115, 146, 151, 180, 181, 198, 208, 215, 218, 220, 231, 235, 257, 268, 269, 276, 282, 381, 417. (2) 23, 59, 76, 141 , 256, 280, 288, 418, 420, 424, 426, 429, 433, 441, 453, 463-465 Carlyle, J. J., (2) 378, 3 8 7 Carnot, C. G., (1) 233, 278. (2) 81, 130 Carpenter, G. S., (2) 416, 453 Carpenter, P. A., (1) 107, 151 Carpenter, S., (1) 73, 96 Carrell, S. E ., (1) 9, 12, 32 Carson, T. P., (1) 380, 406 Carter, T ., (2) 416, 465 Carver, C. S., (1) 133, 146, 327, 332, 367, 406. (2) 167, 180, 263, 265, 277, 280, 286, 299, 329 Castellan, N. J., (2) 429, 443, 456 Catalan, J., (2) 437, 454 Cate, R. M ., (2) 197, 232, 235

Catrambone, R., (2) 302, 303, 324, 329 Cattell, R. B., (1) 364, 406 Cattin, P., (2) 425, 453 Cavazos, L ., (2) 249, 253, 275, 280 Cech, C. G., (1) 177, 206 Cecil-Pigo, E. F., (2) 197, 236 Chaiken, S., (1) 4, 12, 13, 19, 20, 29-31, 32, 34, 39, 124, 125, 128, 129, 131, 132, 135, 145, 146, 265, 266, 276, 392, 407. (2) 70, 72, 80, 81, 83-86, 88-93, 95, 102, 104-107, 111-113, 121, 123-125, 284, 389, 460,

250, 258, 382, 385, 433, 453, Chaires, W. M .,

129, 130, 132, 136, 286, 344, 376, 377,

402, 411-413, 431, 462

(1) 136, 147. (2) 167, 182

Chakravarti, D., (2) 399, 417, 418, 427, 430, 452, 454

Challis, B. H ., (1) 184, 2 0 7 Chambers, W ., (2) 167, 180 Chambless, D. L ., (2) 297, 329 Chan, Y., (1) 173, 198 Chandler, C. C ., (1) 336, 344, 416 Chandrashekaran, M ., (2) 400, 416, 417, 458 Chanowitz, B., (1) 2, 7, 33, 37. (2) 403, 453 Chao, G. T ., (2) 360, 381, 392 Chapman, D ., (1) 133, 144 Chapman, J. P., (1) 164, 187, 188, 196 Chapman, L. S., (1) 164, 187, 188, 196. (2) 12, 59 Chappell, H. W ., (2) 149, 180 Chase, W. G., (1) 113, 146 Chassein, B., (1) 294, 310, 320, 321 Chassin, L., (1) 162, 196 Chattopadhyay, A., (2) 83, 84, 101, 123, 132, 416, 422, 423, 451, 453 Cheever, J., (1) 172, 203 Chen, H. C., (1) 95, 123 Chen, J. A., (2) 346, 366, 370, 371, 393, 448, 454 Cheng, P. W ., (1) 263, 276 Chevron, E. S., (2) 292, 304, 328 Chew, B. R., (1) 157, 200 Chi, M ., (1) 274, 276, 384, 388 Child, P., (1) 170, 203 Childers, T. L ., (2) 407, 408, 454, 457 Chilstron, J. T ., (2) 103, 722 Chrisman, K., (2) 200, 201, 232 Christensen, A., (2) 205, 206, 235 Christensen-Szalanski, J. J. J., (2) 352, 388 Christianson, A. J., (2) 266, 277, 280, 286 Chuang, Y. C ., (2) 81, 130 Cialdini, R. B., (2) 87, 89, 107, 123, 138, 211, 212, 219, 232, 341, 388, 427, 437, 454 Cioffi, D., (2) 243-247, 249-251, 272, 281 Clark, D. A., (2) 297, 329 Clark, D. M ., (1) 380, 406 Clark, E. V., (1) 290, 3 1 7 Clark, H. H., (1) 71, 93, 290-292, 299, 316, 317, 320. (2) 264, 285 Clark, L ., (1) 295, 306, 307, 320 Clark, L. A ., (2) 291, 329 Clark, L. F., (1) 24, 26, 27, 36, 134, 138, 148, 387, 406. (2) 240, 242, 243, 245, 259, 262, 275, 278, 281, 284, 287

AUTHOR INDEX

Clark, M. S., (1) 5, 35, 374, 376, 380, 397, 410. (2) 101, 123, 175, 182, 196, 198-200, 208, 217-219, 232, 233, 235, 238

Clark, R. D., (2) 102, 103, 131 Clark, S. E., (1) 139, 140, 148 Clary, E. G., (2) 35, 59 Clements, C. M ., (2) 294-296, 320, 3 2 7 Clemons, D. S., (2) 91, 137, 405, 411, 463 Clemons, T ., (2) 368, 392 Cleveland, J. N., (2) 344, 368, 388 Cleveland, S. E ., (1) 187, 197 Cloninger, C. R., (2) 291, 334 Clore, G. L., (1) 11, 13, 14, 38, 324-329, 332-339, 341, 342, 345, 348, 352, 359, 361-363, 366-369, 380-386, 388, 389, 404, 4 06-408, 412, 414. (2) 175, 186, 218, 237, 262, 286 Coe, C. L ., (2) 258, 286 Coffman, T. L., (2) 49, 62 Cohen, B. D., (1) 167, 168,197,205 Cohen, B. M ., (1) 256, 275 Cohen, C. E ., (1) 90, 93. (2)33,34,59 Cohen, F., (2) 248, 281 Cohen, G., (1) 189, 196 Cohen, J. B., (2) 399, 454 Cohen, J. D., (1) 135, 146 Cohen, O., (1) 27, 39, 40 Cohen, P. R., (1) 399, 406 Cohen, R., (1) 373, 412 Cohen, R. F., (2) 270, 271, 281 Cohen, S., (2) 219, 233, 251, 281 Cole, E., (1) 223, 283. (2) 35, 66 Cole, S. W ., (1) 8, 31, 101, 144, 246, 258, 275. (2) 20, 58, 325, 32 7 Coley, J. D., (1) 227, 277 Collins, A. M ., (1) 44, 93, 332-339, 341, 345, 348, 359, 361-363, 367-369, 404, 412. (2) 164, 168, 181, 256, 281, 418, 454 Collins, B. E., (2) 73, 81, 130 Collins, J. C., (1) 142, 147 Collins, J. E., (1) 72, 98, 387, 406. (2) 213, 238 Collins, N. L., (2) 194, 196, 233 Collins, R. L ., (2) 272, 277, 280, 287, 291, 329 Combs, A. W ., (2) 302, 336 Conlon, E. J., (2) 347, 386 Conover, P., (2) 163, 176, 181 Converse, P. E ., (2) 163, 165, 180, 181 Converse, S. A., (2) 254, 255, 280 Conville, R. L., (2) 77, 124

473

Conway, M ., (1) 48, 313, 317, 319, 370, 371, 394, 395, 400, 407, 409, 415. (2) 115, 137, 262 Cook, H ., (2) 270, 282 Cook, J., (2) 94, 140 Cook, T. D., (1) 265, 278. (2) 81, 83, 113, 114, 121, 124 Cooper, J., (1) 126, 146, 229, 284. (2) 33,68, 95, 96, 105, 112,124-126, 137, 142 Cooper, L. A., (1) 186, 189, 206 Cooper, M. L., (2) 291, 329 Cooper, W. H., (2) 346, 355, 359, 388 Cooper, Z ., (1) 371, 416 Coovert, M. D., (1) 213, 276 Corbett, A. J., (2) 408, 454 Corcoran, D., (2) 199, 200, 232 Cornell, D. P., (2) 96, 139, 213, 230, 238 Corty, E., (1) 60, 97, 162, 196, 254, 258-260, 265, 282, 306, 320. (2) 410, 420, 463 Cosmides, L., (1) 402, 403, 407, 416 Costa, P. T ., (2) 266, 267, 277, 281, 285 Cousins, S., (1) 175, 196 Cowan, C., (2) 95, 139, 225, 238 Cowan, W. B., (1) 3, 16, 31, 36, 115, 148 Cox, A. D., (2) 93, 124 Cox, D. J., (2) 243, 282 Cox, D. S., (2) 93, 124 Cox, M. G., (1) 70, 93. (2) 324, 330 Coyne, C., (1) 371, 413 Coyne, J. C., (1) 290, 299, 300, 309, 318, 329, 33 7

Craik, F. I. M ., (1) 54, 66, 93, 176, 196 Cramer, P., (1) 379, 407 Crandall, C. J., (2) 297, 329 Crandall, J. E ., (2) 217, 233 Crane, M ., (1) 166, 196, 202 Crano, W. D., (2) 81, 105, 117, 124, 138, 408, 463 Cranford, J. A., (1) 154, 195 Craske, M. G., (2) 297, 301, 329 Crawford, H. L ., (2) 108, 138 Crelia, R., (1) 12, 28, 36, 54, 95, 127, 131, 148. (2) 168, 183 Crews, J., (2) 244, 282 Creyer, E. H ., (2) 425, 430, 462 Crites, S., (2) 71, 124 Crocker, J., (1) 2, 9, 23, 33, 39, 40, 58, 93, 179, 207, 230, 253, 254, 281, 284. (2) 34, 47, 51-53, 59, 64, 67, 112, 124, 233, 293, 306, 307, 327, 33 7

474

AUTHOR INDEX

Crockett, W. H ., (1) 167, 196 Cronbach, L. J., (2) 353, 355, 379, 388 Cronen, V. E., 77, 124 Cross, S. E., (1) 154, 202 Crovitz, H. F., (1) 193, 196 Crow, K., (1) 393, 405 Crowley, N., (2) 18, 59 Croyle, R. T ., (2) 248, 249, 251, 254, 255, 261, 268, 269, 275, 281, 283, 286 Crum, R., (2) 152, 185 Crusco, A. H., (1) 11, 33 Csank, P., (1) 3 1 1 ,4 0 7 Cullen, J., (2) 244, 282 Cunniff, C., (1) 13, 26, 40, 51, 52, 98, 267, 284

Cunningham, M ., (2) 218, 219, 233 Cunningham, R. L ., (1) 171, 200 Cutrona, C. E ., (2) 291, 329 Cutting, J. E., (1) 186, 201 Czyzewska, M ., (1) 101, 147, 223, 283

Deaux, K., (1) 21, 33. (2) 18, 59, 362, 363, 388, 389

Debner, J. A., (1) 14, 35 DeBono, K. G., (1) 137, 145. (2) 73, 74, 99, 124, 138

Deci, E. C., (2) 105, 124 Deckel, A. W ., (2) 244, 283 Decker, S., (2) 270, 286, 291, 3 36 de Houwer, J., (1) 19, 35 DelBoca, F. K., (2) 2, 15, 58 Deighton, J., (2) 401, 421, 442-444, 454, 4 57

Dekleva, C., (2) 430, 431, 464 De La Ronde, C., (2) 291, 3 3 7 Dellarosa, D., (2) 444, 454 DeLongis, A., (2) 291, 330 DeMaio, T. J., (1) 289, 31 7 DeMayo, R., (2) 300, 331 Dember, W. N., (2) 289, 329 Dembroski, T. M ., (2) 261, 267, 277, 281, 285

D DAgostino, P. R., (1) 8, 11, 24, 29, 30, 33, 37, 319, 413. (2) 110, 121 Dakof, G. A., (2) 270, 272, 277, 280, 281, 287

Dalgleish, T ., (2) 300, 301, 329 DAndrade, R., (2) 80, 140, 359, 395 Damasio, A. R., (1) 187, 196 Damasio, H., (1) 187, 196 Damon, W ., (1) 171, 173, 174, 196, 198 Danilovics, P., (1) 395, 408. (2) 319, 330 Darby, B. L., (2) 219, 232, 437, 454 Darley, J. M ., (1) 12, 33, 221, 230, 276. (2) 33 , 38, 59, 248, 249, 282, 322, 331, 360, 362, 368, 377, 388, 443, 454 Darwin, C., (1) 330, 407 Daubman, K. A., (1) 396, 397, 409. (2) 101, 128

Davidson, A. R., (2) 81, 111, 116, 124 Davis, K. E., (1) 87, 95, 218, 279. (2) 104, 127, 220, 233, 363, 391 Davis, L. J., (1) 188, 208 Davis-Blake, A., (2) 342, 388 Davison, G. C., (2) 290, 331 Davitz, J. R., (1) 330, 390, 407 Dawes, R „ (1) 135, 149, 232, 276. (2) 290, 329, 410, 454 Day, K. D., (1) 385, 4 17 Dearing, J. W ., (2) 155, 185

Demerath, N. J., (2) 70, 132 DeNicholas, M. E., (2) 211, 232 DeNisi, A. S., (2) 350-352, 372, 374, 375, 385, 387, 388, 396, 3 9 7 DePaulo, B. M ., (1) 70, 76, 93 DePauw, S., (2) 244, 284 DeRivera, J., (1) 330, 4 07 Dermer, M ., (1) 188, 204 Derry, P. A., (1) 175, 201, 379, 407. (2) 297, 304, 329, 333 DeRubeis, R. J., (2) 299, 311, 3 26 DeSarbo, W. S., (2) 93, 122, 406, 461 Desmarais, S., (2) 197, 233 Detweiler, M ., (1) 118, 149 Deutsch, B., (1) 314, 320 Deutsch, M ., (2) 198, 233 DeVellis, R., (2) 270, 282 Devine, P. G., (1) 4, 8, 10, 11, 13, 21, 22, 28, 29, 33, 60, 73, 93, 256, 263, 271, 276. (2) 7, 18, 24, 39, 40, 42, 46, 51, 52, 59, 70, 97, 121, 124, 341, 344, 351, 365, 368, 376, 377, 382, 388 Dewey, G. J., (1) 139, 148 Dick, A., (2) 427, 430, 454 Dickson, P. R., (2) 91, 133, 401, 465 DiClemente, C. C ., (2) 291, 329 Diefenbach, M ., (2) 254, 284 Diehl, M ., (2) 8, 9, 59, 96, 139 Diener, C. I., (1) 358, 407 Dienes, B., (1) 326, 406

AUTHOR INDEX

Dijker, A. G. M ., (2) 46, 59 Dillehay, R. C., (2) 77, 124 Dion, K., (1) 218, 276 Dipboye, R. L., (2) 349, 369, 388 Ditto, P. H ., (2) 248, 249, 269, 281-283 Dittus, P., (2) 117, 128 Dixon, J. C ., (1) 172, 196 Dixon, T. M ., (1) 167, 196 Dobbins, G. H., (2) 359, 364, 377, 388 Dobson, D. J. G., (1) 395, 399, 4 07 Dobson, K. S., (1) 395, 399, 407. (2) 297,

475

Dunkel-Schetter, C ., (2) 291, 330 Dunn, D. S., (2) 72, 81, 117, 140, 141 Dunnette, M. D., (2) 340, 388 Dunning, D., (1) 233, 276, 278. (2) 427, 456 Durso, F. T ., (2) 250, 285 Dutton, D. G., (2) 221, 233 Duval, S., (2) 298, 329 Dworkin, R. H., (1) 166, 196 Dyck, M. J., (2) 296, 297, 299, 301, 308, 331

Dyer, L. M ., (1) 21, 34

329

Dodd, D. H ., (1) 298, 3 1 7 Dodge, K. A., (2) 51, 60 Doherty, M. E., (2) 384, 3 8 6 , 444, 454 Donohue, T ., (2) 160, 186, 187 Doren, B., (1) 380, 409 Dorfman, D., (1) 16, 35 Dornoff, R. J., (2) 400, 416, 417, 458 Dosher, B. A., (2) 408, 454 Dovidio, J. F., (1) 6, 9, 33, 37, 256, 271, 276. (2) 39, 45, 59 Dowd, T. G., (2) 95, 133 Downing, J. W ., (1) 137, 142, 147. (2) 77, 129

Downs, A., (2) 165, 181 Downs, A. C., (2) 368, 388 Drake, C. A., (2) 262, 275, 285 Drake, R. A., (1) 142, 147. (2) 77, 129 Dreben, E. K., (1) 56, 93 Driscoll, D. M ., (1) 14, 16, 38, 58, 59, 94, 180, 198, 219, 282. (2) 15, 36, 60, 61, 12, 73, 125, 138, 168, 181 Druian, P., (1) 84, 9 6 Duan, C., (1) 227, 234, 269, 283. (2) 33, 35, 36, 66 Dubois, C. L. Z., (2) 346, 363, 366, 395 Duda, K., (2) 291, 328 Dugan, P. M ., (1) 213, 278. (2) 13, 61 Duhe, A., (2) 222, 223, 231 Dukerich, J. M ., (1) 24, 3 7 Dulany, D. E., (1) 308, 3 1 7 Dull, V. T ., (1) 251, 276. (2) 7, 17, 18, 50-52, 59 Dumais, S. T., (1) 105, 114, 150 Dunbar, K., (1) 137, 146 Dunbar, S., (2) 23, 24, 62 Duncan, B. C., (2) 33, 59 Duncan, C. P., (2) 401, 454 Duncan, S. L., (1) 229, 276 Duncker, K., (1) 397, 407

E Eagly, A., (1) 124, 125, 127, 131, 132, 135, 146, 246, 265, 276, 392, 407. (2) 80, 82, 83, 85, 86, 88, 91, 93, 94, 105-107, 111, 123-125, 129, 344, 346, 361, 363, 367, 376, 377, 382, 385, 389, 391, 411, 413, 453, 457 Easterbrook, J. A ., (1) 392, 406 Eaves, L. J., (2) 76, 125 Ebbesen, E. B., (1) 42, 78, 79, 90, 9 2 -9 4 , 180, 198, 374, 412 Ebbinghaus, H., (1) 52, 93 Edelstein, W ., (1) 174, 198 Eden, D., (2) 357, 389 Eder, R. A., (1) 172, 196 Edwards, J. A., (1) 385, 391, 392, 394, 395, 407, 416. (2) 295, 317-320, 329, 338 Edwards, K., (1) 9, 11, 12, 33, 251, 280, 385, 410. (2) 74, 125 Eelen, P., (1) 19, 35 Effrein, E. A., (2) 23, 60 Egeland, B., (2) 192, 233 Egeth, H. E., (1) 3, 34 Eich, E., (2) 261, 275, 282 Eich, J. E., (1) 375, 407 Eidelson, J. I., (2) 297, 327 Einhom, H. J., (2) 377, 389, 421, 457 Eisdorfer, C., (2) 200, 280 Eisenman, E. J., (2) 370, 393 Eiser, J. R., (1) 288, 3 1 7 Ekman, P., (1) 327-331, 390, 407, 411 Elbin, S. D., (2) 305, 322, 338 Elkes, R., (2) 437, 453 Elkin, R. A., (2) 91, 96, 116, 125, 130 Elliott, A. J., (1) 21, 33. (2) 46, 59 Ellis, H. C., (1) 370-372, 375, 388, 400, 401, 407, 408, 410. (2) 418, 422, 465

476

AUTHOR INDEX

Ellsworth, P. C ., (1) 298, 320 , 329-331, 336, 338, 343, 344, 348-352, 354, 363, 365, 385, 404, 407, 408, 410, 415 Elmore-Yalch, R., (2) 92, 141 , 402, 412, 466 Elms, A. C., (2) 87, 113, 125 Emery, G., (1) 114, 145. (2) 293, 297, 298, 328

Emmons, R. A., (1) 335, 408 Enelow, J., (2) 153, 165, 181 Engle, R. W ., (2) 108, 138, 139 , 412, 463, 464

Engquist, C., (2) 243, 245, 284 Engquist, G., (2) 370, 394 Epstein, N., (2) 297, 328 Epstein, R., (1) 172, 196 Epstein, S., (1) 164, 165, 196, 231, 266, 267, 276, 277. (2) 291, 3 30 Erb, H. P., (1) 393, 405 Erbaugh, J., (2) 313, 328 Erber, R., (1) 8, 29, 33, 252, 276. (2) 102, 125, 160, 183, 216, 217, 232, 233 Erbring, L., (2) 156, 181 Erdelyi, M. H ., (1) 10, 33. (2) 404, 422, 440, 454 Erdley, C. A., (1) 8, 11, 33 Erickson, G. A., (2) 211, 235 Erickson, M ., (2) 192, 233 Ericsson, K. A., (1) 113, 146. (2) 417, 454 Erikson, E. H., (1) 128, 146 Ernst, D., (2) 296, 297, 301, 308, 331 Ernst, J. M ., (2) 72, 121 Eron, L. D., (2) 87, 128 Esses, V. M ., (2) 46, 60 Estes, W. K., (1) 180, 197 Etcoff, N. L., (1) 227, 283. (2) 9, 10, 6 7 Evans, F. J., (1) 190, 197, 200 Evans, M ., (1) 76, 96. (2) 34, 65, 443, 462 Evans, M. G., (2) 77, 125 Evans, N., (1) 6, 9, 33, 256, 271, 276. (2) 39, 59 Evans, T. W ., (1) 383, 411. (2) 102, 132 Evenbeck, S., (2) 427, 454 Evev, I., (2) 194, 235 Ewort, O., (1) 384, 408 Eysenck, H. J., (2) 7 0 , 125, 292, 301, 3 0 2 , 330, 334

Eysenck, M. W ., (1) 392, 408

F Fabrigar, L ., (2) 71, 107, 121, 124, 135 Falke, R. L ., (2) 270, 2 8 7

Fallon, A. E ., (1) 188, 197, 206, 342, 413 Farah, M. J., (1) 186, 197 Farquhar, P. H ., (2) 431, 454 Farr, J. L ., (2) 345-347, 385, 392 Farrell, D ., (2) 203, 236 Faust, D., (1) 232, 276 Favero, J. L ., (2) 342, 362, 366, 391 Fawzy, F. I., (2) 244, 248, 285 Fazio, R. H ., (1) 4, 5, 9, 12, 16,18, 20, 33, 35, 38, 119, 125, 126, 136, 142,144, 146, 149, 219, 231, 264, 277, 282. (2) 23, 60, 70, 72, 80, 81, 95, 96, 104, 105, 107, 112, 117, 118, 121, 124-126, 128, 129, 136-138, 155, 167, 181, 182, 224, 237, 250, 282, 322, 330, 344, 389, 405, 408, 428, 431, 437, 443, 448-450, 454-456, 462

Fenster, D., (2) 266, 280 Feeney, J., (2) 194, 233 Fegley, S., (1) 167, 173, 198 Fehr, B., (1) 324, 359, 408. (2) 309, 328 Fein, S., (1) 230, 277 Feldman, J. M ., (2) 83, 126, 340-343, 345, 347-351, 355-357, 361-364, 368, 369, 371, 374, 376, 378-380, 382, 383, 385, 389, 319, 397, 438, 441, 455

Feldman, S., (2) 155, 163, 176, 181, 187 Fernandez, G., (1) 395, 408. (2) 319, 330 Festinger, L ., (1) 10, 34, 127, 146, 303, 314, 317. (2) 95-97, 126, 209, 210, 233, 267, 282, 302, 321, 330 Fetter, R., (2) 345, 358, 393 Fetzer, B. K., (2) 271, 285 Fiedler, J., (1) 302, 319 Fiedler, K., (1) 308, 317, 372, 380, 382, 388, 398-400, 408. (2) 14, 60, 217, 233, 319, 33 6 Fifield, J., (2) 270, 271, 279 Fincham, F. D ., (2) 222-225, 2 3 2-234, 290, 291, 330 Fincher-Kiefer, R., (1) 29, 33 Findlay, B., (2) 244, 284 Fiorina, M. P., (2) 147, 149, 152, 181 Firth-Cozens, J., (2) 290, 330 Fischer, G. W ., (1) 257, 279. (2) 18, 21, 26-28, 63, 250, 264, 274, 278, 284 Fischer, P., (2) 87, 128 Fischoff, B., (1) 135, 146. (2) 154, 186, 250, 251, 260, 264, 274, 278, 284, 424, 429, 443-445, 455, 459, 463 Fish, T. A., (2) 256, 282

AUTHOR INDEX

Fishbein, M ., (2) 77-79, 110, 111, 116, 117, 120, 121 , 164, 167, 172, 173, 181, 185 , 449, 451, 455 Fishbein, S., (2) 217, 221, 238 Fisher, D. L., (2) 100, 136 Fisher, L ., (1) 173, 187, 198 Fisher, R. B., (2) 261, 282 Fisher, S., (1) 187, 197 Fisicaro, S. A., (2) 355, 356, 377, 389, 392 Fiske, A. P., (2) 196, 234 Fiske, S. T ., (1) 2, 4, 7, 11, 13, 14, 16, 21, 28-30, 33, 34, 39, 43, 47, 56, 70, 73, 88, 93, 94, 125, 126, 129-131, 137, 138, 141, 148, 215, 216, 226, 227, 229, 244, 250, 252, 253, 260-262, 269, 277, 283, 304, 321, 394, 408. (2) 7, 9, 10, 16, 17, 21, 24, 39, 40, 42, 46, 60, 64, 67, 71, 89, 109, 112, 119, 120, 124, 126, 160, 161, 164, 176, 179, 181, 183, 196, 225, 234, 236, 240, 246, 252, 256, 282, 292, 293, 302, 318, 319, 324, 330, 343, 344, 348, 349, 351, 352, 362, 363, 377, 389, 395, 396, 402, 405, 412, 429-432, 435, 455 Fiskin, S. A., (2) 262, 275, 385 Fitts, P., (1) 105, 112, 146 Fivush, R., (1) 156, 197 Flament, C., (2) 8, 67 Flavell, J., (1) 172, 179, 197 Flay, B., (2) 83, 113, 114, 124 Fletcher, B., (2) 242-246, 277, 286 Fletcher, J. F., (2) 81, 115, 121 Fletcher, G. J. O., (2) 395, 408. (2) 229, 234, 319, 330 Flett, G. L., (2) 319, 321, 330 Fodor, J. A., (1) 269, 277 Foesterling, F., (2) 290, 330 Fogarty, S. T., (1) 376, 416 Folger, R., (1) 12, 38 Folkman, S., (1) 263, 410. (2) 242, 263, 278, 282, 283, 291, 330 Follette, W. C., (2) 223, 235 Fonagy, P., (2) 196, 234 Fondacaro, R. A., (2) 94, 132 Fong, G. T ., (1) 166, 197, 246, 277. (2) 302, 330, 365, 379, 389 Ford, G. T ., (2) 424, 430, 455 Ford, J. K., (2) 346, 365-367, 375, 390, 392 Ford, M. E., (1) 398, 399, 411 Ford, T. E., (1) 234, 283. (2) 6, 46, 66, 71, 139, 200, 235 Foreyt, J. P., (2) 290, 330

477

Forgas, J. P., (1) 8, 34, 373-375, 379-381, 386, 388, 408. (2) 43, 44, 60, 216, 217, 234 Fomell, C., (2) 416, 426, 462 Forston, M .'T ., (2) 258, 282 Forsyth, D. R., (2) 291, 336 Foss, M ., (1) 325, 406, 412 Foti, R., (2) 340, 373, 375, 383, 384, 390, 393

Fox, J., (1) 107, 108, 146 Francolini, C. M ., (1) 3, 34 Franks, J. J., (1) 121, 122, 148 Fraser, S., (2) 437, 455 Freedman, J., (2) 409, 437, 455 Freedman, M ., (1) 185, 206 Freedman, M. R., (2) 409, 452 Freeman, S., (2) 211, 212, 232 Freize, I. H ., (2) 222, 238, 350, 396 French, R., (2) 291, 3 2 7 Frenkel-Brunswik, E ., (2) 145, 179 Freund, T ., (2) 93, 130 Frey, D., (2) 94, 126 Fridh, G., (2) 261, 262, 285 Fridhander, B. M ., (1) 171, 207 Fried, L. S., (2) 22, 24, 28, 60 Friedman, A., (2) 403 , 455 Friedman, S., (1) 12, 36 Friesen, W. V., (1) 329-331, 407, 411 Frijda, N. H., (1) 326-330, 337, 338, 343-345, 350, 353, 355, 363, 365, 366, 388-391, 402, 404, 408, 412 Froming, W. J., (2) 167, 180 Fuhrman, R. W ., (1) 42, 72, 83, 96, 98, 182, 184, 190, 191, 197, 201, 20 8 , 306, 322. (2) 23, 62, 159, 187 Fujioka, T., (1) 170, 203 Fujita, F., (1) 326, 406 Fulero, S., (1) 74, 76, 96, 214, 216, 260, 281. (2) 5, 22, 24, 34, 65, 443, 462 Funder, D. C., (2) 351, 352, 358, 381, 390 Funk, C. L., (2) 161, 186 Furnham, A., (2) 306, 307, 328, 330 Furse, D. H., (2) 401, 455

G Gabrielcik, A., (1) 9, 34 Gaelick, L ., (1) 178, 207. (2) 224, 234, 302, 336

Gaertner, S. L., (1) 45, 59

478

AUTHOR INDEX

Gaeth, G. J., (2) 446, 459 Galambos, J. A., (1) 76, 92 Gallagher, D., (1) 384, 408 Galley, D. J., (1) 9, 33, 141, 145 Gallup, G. G., (1) 172, 197 Ganellen, R., (2) 167, 180 Gangestad, S. W ., (2) 227, 23 7 Gannett, B. A., (2) 346, 366, 370, 371, 393 Gansler, D. A., (2) 301, 332 Gara, M. A., (1) 167, 168, 171, 197, 205, 206

Garamoni, G. L ., (2) 291, 336 Garber, J., (2) 296, 297, 307, 317, 320, 330, 331

Garcia, M ., (1) 19, 34. (2) 379, 393 Garcia, S., (1) 76, 95 Gardner, M. P., (2) 108, 126 Garrity, T. F., (2) 244, 282 Garrod, S., (1) 291, 31 7 Gartska, T ., (2) 262, 281 Gaschke, Y. N., (1) 383, 411. (2) 102, 132 Gaskin, S., (2) 416, 463 Gaston-Johansson, F., (2) 261, 262, 285 Gaudet, H., (2) 146, 165, 183 Gavanski, I., (2) 410, 455, 463 Geen, R. G., (1) 237, 238, 275 Geller, V., (1) 2, 34 Gelman, S. A., (1) 227, 277 Genero, N., (1) 160, 196 Gent, M ., (2) 244, 284 Gentner, D., (1) 236, 237, 241, 257, 262, 277 Gerard, H. B., (2) 33, 34, 68, 96, 126 George, A. L., (2) 145, 181 George, C., (2) 194, 234 George, J. L., (2) 145, 181 George, J. M ., (2) 265, 280 George, W. H ., (2) 291, 329 Gergen, K. J., (1) 191, 197 Gergen, M. M ., (1) 191, 197 Gersho, B. F., (2) 44, 63 Geva, N., (1) 181, 204. (2) 167, 183 Giannopolous, C ., (1) 371, 394, 40 7 Gibbons, F. X., (2) 291, 321, 331 Gibbons, P., (2) 55, 61 Gibson, B. D., (2) 428, 462 Gibson, F. W ., (2) 380, 391 Gick, M. L ., (1) 236, 277 Gifford, R. K., (1) 212, 278. (2)11-13, 61 Gigenzerer, G., (1) 104, 146, 306,308, 3 1 7 Gijsbers, K., (2) 261, 262, 286

Gilbert, D. T ., (1) 3, 7, 18, 22-24, 26-28, 34, 39, 44, 60, 64, 66, 94, 115, 126, 127, 131, 146, 221, 253, 263, 268-271, 277. (2) 24, 40-42, 60, 19, 119, 126, 139, 344, 348, 351, 352, 368, 376, 390, 421, 455 Gilbert, G. M ., (2) 49, 60 Gillig, P. M ., (2) 114, 126 Gilligan, S. G., (1) 176, 195, 373, 375, 376, 406, 409. (2) 174, 180, 313, 328 Gillund, G., (1) 177, 197 Gilovich, T ., (1) 8, 34, 239, 277. (2) 20, 60, 444, 455 Giner-Sorolla, R., (1) 19, 34. (2) 72, 84, 123 Gioia, D. A., (2) 342, 345, 362, 372, 377, 392

Glaser, R., (2) 384, 388 Glass, B., (1) 227, 269, 283 Gleicher, F., (1) 391, 392, 395, 416. (2) 74, 100, 102, 126, 135, 295, 317-320, 329, 331, 338, 406, 455 Gleitman, H ., (1) 163, 195 Gleitman, L. R., (1) 163, 195 Glick, P., (1) 230, 277. (2) 364, 390 Godden, D. R., (1) 373, 375, 409 Godfrey, S. S., (1) 86, 92 Goethals, G. R., (2) 322, 331 Goetz, E. T ., (2) 166, 180 Gohm, C ., (1) 361, 409 Goldberg, L ., (2) 76, 141 Goldberg, L. R., (1) 161, 197. (2) 55, 61 Goldberg, M. E ., (2) 93, 101, 722, 126 Goldenberg, E., (2) 156, 181 Goldsamt, L. A., (1) 371, 416 Goldstein, D., (1) 142, 147. (2) 409, 456 Goldstone, R., (1) 242, 262, 272, 278 Goleman, D ., (1) 12, 34 Golin, S., (2) 310, 33 7 Gollwitzer, P., (1) 16, 19, 31, 32, 34, 133, 134, 14 7 Gonder-Frederick, L. A ., (2) 243, 282 Gonzales, M. H ., (1) 309, 311, 3 1 7 Goodkin, K., (2) 265, 280 Goodnow, J. J., (1) 158, 195 Goodyear, M. D ., (2) 244, 284 Gordon, P. C ., (1) 9, 34, 177, 178, 199. (2) 302, 332 Gordon, S. E ., (1) 47, 52, 55-57, 77, 80, 94, 98, 180, 208. (2) 167, 168, 187, 250, 266, 288, 433, 466 Goreczny, A. J., (1) 188, 208

AUTHOR INDEX Gorenflo, D. W ., (2) 105, 124 Gom, G. J., (2) 93, 100, 101, 126 , 412, 455 Gorsuch, R. L ., (2) 313, 320, 336 Gossop, M ., (2) 291, 328 Gotlib, I. H ., (2) 296, 298-300, 309, 329, 331

Gottman, J., (2) 229, 234 Gotz, A., (2) 422, 440, 455 Gouaux, C ., (2) 108, 126 Govender, R., (1) 4, 12, 19, 20, 31, 32, 125, 145. (2) 70, 212 Graber, D ., (2) 160, 181 Graen, G., (2) 340, 357, 365, 377, 390 Graesser, A. C ., (1) 77, 80, 94, 191, 204 Graf, P., (1) 6, 37, 176, 206. (2) 87, 138 Graham, S., (1) 329, 336, 344, 416 Graham-Pole, J., (2) 263, 286 Granbois, D. H., 401, 461 Grandits, G. A., (2) 266, 281 Gray, J. A., (1) 329, 409 Gray, W ., (1) 159, 162, 205 Green, G. M ., (1) 71, 72, 94 Green, L., (2) 291, 328 Green, S. G., (2) 347, 393 Greenberg, J., (1) 12, 38. (2) 198, 201, 234, 262, 285, 291, 299, 301, 307, 321, 333, 335, 336

Greenberg, M. S., (2) 297, 312, 314, 315, 331

Greenberger, D., (2) 197, 238 Greene, D., (2) 105, 130, 306, 335 Greenwald, A. G., (1) 9, 10, 12, 20, 34, 42, 59, 94, 154, 169, 175-177, 197, 207, 233, 278. (2) 41, 58, 71, 80, 85-87, 91, 97, 112-114, 126, 127, 131, 135, 136, 292, 331, 402, 411, 414, 416, 455, 456, 462

Gresham, L. G., (2) 108, 127 Grether, D., (2) 152, 181 Grice, H. P., (1) 71, 94, 290, 292, 299, 302, 318. (2) 79, 127, 373, 376, 390 Griffeth, R. W ., (2) 341, 393 Griffin, D. W ., (1) 233, 276, 278, 284. (2) 70, 72, 140, 424, 427, 456 Griffiths, R. J., (2) 35, 36, 63 Griffitt, W. B., (2) 100, 127, 217, 234 Gross, P. H., (1) 12, 33, 221, 230, 276. (2) 33, 3 8 , 59, 360, 362, 368, 377, 3 8 8 , 443, 454 Gross, S., (2) 81, 127 Grader, C. L., (1) 265, 278. (2) 81, 113, 114, 121, 124

47 9

Graen, R., (2) 291, 330 Graenfeld, D. H ., (1) 58, 66, 71, 94, 98. (2) 79, 127 Grash, J. E., (2) 110, 127 Gratzinger, R., (2) 249, 253, 275, 280 Gschneidinger, E ., (1) 387, 415 Gurin, G., (2) 148, 165, 180 Gurtman, M. B., (1) 6, 9, 20-22, 3 7 Gurumurthy, K., (2) 400, 416, 417, 425, 426, 456, 458 Gurwitz, S. B., (2) 51, 60 Gutman, M ., (2) 117, 128 Guyatt, G. H ., (2) 244, 284 Guzzo, R. A., (2) 381, 393

H Ha, Y., (1) 132, 134, 147. (2) 348, 392, 428, 442-444, 456-458 Haaga, D. A. F., (2) 290, 296, 297, 299, 301, 308, 331 Haberfield, Y., (2) 363, 394 Haberkom, G., (2) 103, 131 Hackett, C., (2) 409, 452 Hackett, T ., (2) 244, 282 Haddock, G., (2) 46, 60 Hager, W ., (1) 380, 411 Hagerty, M. R., (2) 407, 451 Hailey, B. J., (2) 267, 283 Halal, M ., (1) 377, 410 Haley, J. V., (1) 168, 199 Hamill, R., (1) 245, 278. (2) 166, 167, 169, 183

Hamilos, C. A., (1) 260, 278 Hamilton, D. L ., (1) 24, 34, 42, 51, 52, 55, 56, 58, 59, 61, 63, 74, 94, 180, 197, 198, 211-215, 222-226, 235, 260, 278, 282, 283. (2) 2-4, 11-15, 19, 22, 23, 31-36, 42-45, 52, 55, 60, 61, 63, 65, 67, 168, 181, 302, 331, 366, 390

Hamilton, J., (2) 299, 301, 335 Hammen, C. L ., (2) 206, 234, 300, 331 Hamond, N. R., (1) 156, 197 Hampson, S. E ., (2) 55, 61 Hand, H. H., (2) 341, 393 Hannah, D. B., (1) 58, 93. (2) 114, 127, 436, 451 Hansen, C. H., (1) 5, 16, 35 Hansen, R. D., (1) 5, 16, 35

480

AUTHOR INDEX

Hanson, C ., (1) 260, 278 Hantas, M ., (1) 376, 377, 412 Harackiewicz, J. M ., (1) 193, 200 Hardy, J. D ., (2) 266, 277, 282 Harkins, S. G ., (1) 54, 96. (2) 90, 91, 727, 135

Harlow, T. F., (1) 385, 411 Ham er, D. E ., (1) 188, 208 Hamish, R. J., (2) 74, 124 Harring, K., (2) 270, 282 Harris, H. J., (2) 103, 130 Harris, R. J., (2) 408, 410, 456 Harris, T ., (2) 292, 328 Harris, V. A ., (1) 126, 147 Hart, D., (1) 167, 172, 173, 196, 198 Harter, S., (1) 172, 198 Hartlage, S., (2) 296, 3 2 7 Hartman, K. A., (2) 253, 254, 283 Hartshome, H ., (1) 161, 198 Hartup, W ., (1) 172, 199 Hartwick, J., (1) 66, 98. (2) 76, 141 , 155, 164, 167, 187, 250, 288, 449, 463 Harvey, J. H ., (2) 205, 206, 235, 290, 291, 332, 338

Harvey, J. S., (2) 319, 328 Harvey, R., (2) 346, 390 Hasher, L ., (1) 2, 35, 86, 92, 142, 147, 370, 380, 409. (2) 409, 456 Hashtroudi, S., (1) 119, 150 Haslam, N ., (2) 196, 234 Hastak, M ., (2) 411, 456 Hastie, R., (1) 24, 35, 42, 51, 56, 57, 59-61, 65, 85, 86, 93, 94, 96, 154, 179-181, 191, 198, 200, 205, 209, 211, 215, 221, 223, 241, 254, 260, 278, 281, 373, 381, 409. (2) 18, 22, 24, 25, 28-31, 34, 61, 62, 64, 83, 127, 161, 168, 176, 181, 185, 246, 250, 252, 266, 282, 319, 331, 373, 379, 390, 394, 406, 420-423, 430, 433, 436, 456 Hatfield, E ., (2) 197, 234, 238 Hauenstein, N. M. A., (2) 380, 381, 383, 384, 390 Haugen, J. A., (1) 311, 31 7 Haugtvedt, C. P., (2) 84, 91, 92, 99, 104, 107, 112, 113, 115, 127, 134, 135 Hauser, J. R., (2) 400, 415, 416, 456 Hausknecht, D., (2) 99, 100, 133, 402, 412, 460

Hautzinger, M ., (2) 299, 333

Haviland, S., (1) 71, 93, 292, 299, 316, 31 7 Hawkins, S. A ., (2) 79, 127, 409, 456 Hay, J., (2) 197, 234 Hayes-Roth, B., (1) 21, 35, 274, 278. (2) 16, 61

Hayes-Roth, F., (2) 16, 61 Hayman, C. A. G ., (1) 119, 150 Hazan, C ., (2) 193, 195, 234, 2 3 7 Head, H ., (1) 187, 198 Hearst, E ., (1) 219, 280. (2) 428, 429, 451, 461

Heath, L ., (2) 256, 282 Heckhausen, H ., (1) 19, 34 Heckler, S. E ., (2) 407, 408, 454, 457 Hedge, J., (2) 383, 384, 390 Hedges, L ., (1) 191, 199 Heesacker, M. H ., (2) 85, 86, 93, 99, 127, 134

Heider, F., (1) 42, 94, 310, 318, 356, 409. (2) 95, 97, 127, 211, 234, 302, 331 Heider, K., (1) 329, 411 Heikkila, K., (2) 267, 283 Heilman, M ., (2) 38, 61, 362, 364, 365, 367, 389-391

Heimberg, R. G., (2) 301, 332 Heise, D ., (1) 325, 412 Helbig, D. W ., (2) 117, 128 Helgeson, V. S., (2) 228, 234 Heller, K., (2) 217, 236 Helmholtz, H ., (1) 303, 318 Helson, H., (1) 288, 318 Hemenway, K., (1) 162, 2 0 7 Henik, A., (1) 3, 35 Hennigan, K. M ., (1) 265, 278. (2) 113, 114, 124

Henninger, M ., (2) 77, 127 Henry, S. M ., (1) 387, 406. (2) 259, 281 Henton, J. M ., (2) 197, 232, 235 Hepburn, C., (1) 253, 254, 279. (2) 38, 39, 63, 361, 392 Herbert, M ., (2) 228, 237, 264, 28 7 Herdt, G., (1) 174, 2 0 7 Herek, G. M ., (2) 73, 127 Herlihy, J. M ., (2) 365, 390 Herman, C. P., (1) 42, 94. (2) 427, 454 Hermans, D ., (1) 19, 35 Herr, B. M ., (2) 346, 362, 366, 370, 371, 393 Herr, P. M ., (1) 5, 12, 35, 54, 94, 219, 264, 277. (2) 70, 125, 167, 182, 250, 282, 405, 407, 418, 420, 423, 427, 428, 431, 440, 441, 443, 448, 454-456, 458, 462

AUTHOR INDEX

Herstein, J. A., (1) 24, 37. (2) 173, 174, 178, 182 Hertel, P. T ., (1) 372, 388, 401, 409. (2) 418, 422, 465 Hervitz, E. F., (2) 423, 463 Hesse, F. W ., (1) 374, 415 Hewstone, M ., (2) 35, 36, 42, 51, 52, 54, 61-63

Hibbert, G. A., (2) 297, 331 Hibbs, D. A., (2) 147, 148, 182 Higgins, E. T ., (1) 2, 5, 6, 8, 11, 12, 14, 16, 23, 24, 29, 30, 32, 35, 36, 39, 42, 44, 53, 54, 60, 66, 71, 94, 95, 104, 116, 117, 119, 127, 129, 132, 136, 137, 143, 144, 147, 150, 163, 169, 172, 198, 199, 219, 228, 246, 256, 260, 261, 278, 279, 292, 300, 305, 309-311, 313, 318, 319, 335, 338, 347, 359-362, 365, 368, 381, 390, 404, 409, 412, 415. (2) 8, 40, 41, 61, 62, 80, 94, 128, 132, 154, 161, 166, 167, 182, 258, 266, 283, 286, 295, 297, 305, 331, 336, 341, 391, 416, 431, 443, 456 Highhouse, S., (2) 354, 381, 385, 39 7 Hildebrand-Saints, L ., (1) 395, 409. (2) 318, 331

Hill, M. G., (2) 290, 318, 332, 338 Hill, T ., (1) 101, 147 Hilterman, R. J., (2) 348, 361-364, 378, 389 Hilton, D. J., (1) 303, 304, 306, 308, 317, 318, 320

Hilton, J. L., (1) 212, 215, 224, 230, 277, 279 Himmelfarb, S., (2) 111, 128 Hinich, M ., (2) 153, 165, 181 Hinton, G. E., (1) 111, 149 Hintzman, D. L ., (1) 102, 147, 163, 199. (2) 13, 20, 21, 62, 254, 283 Hippier, H. J., (1) 295, 297, 306, 307, 314, 318, 320

Hirt, E. R., (1) 396, 397, 412. (2) 87, 101, 133, 138, 211, 235, 423, 429, 443, 456, 463

Hixon, J. G., (1) 18, 22, 23, 34, 270, 271, 277. (2) 24, 40-42, 60, 119, 126, 139, 291, 33 7 Hoberman, H., (2) 299, 333 Hobson, C. J., (2) 380, 391 Hoch, S. J., (2) 79, 127, 400, 409, 421, 423, 442-444, 456, 457

481

Hodges, S. D ., (2) 70, 140 Hoehn-Hyde, D ., (2) 310, 332 Hoffman, H. G ., (1) 298, 319 Hofstadter, D., (1) 236, 237, 262, 279 Hogarth, R. M ., (2) 377, 389, 421, 45 7 Holbrook, M. B., (2) 407, 461 Holland, J. H ., (2) 427, 433, 45 7 Hollon, S. D ., (1) 168, 200. (2) 297, 307, 330, 332

Holloway, J., (2) 247, 272, 281 Holloway, W ., (2) 297, 332 Holman, R. H., (1) 162, 206 Holmes, J. G., (1) 12, 32. (2) 196, 198, 226, 228, 235, 2 3 6 Holstein, C., (1) 226, 267, 277 Holt, K., (2) 270, 282, 307, 332 Holt, L. E ., (2) 113, 140 Holtgraves, T ., (1) 69, 90, 95, 309, 310, 318 Holtz, R., (2) 81, 127 Holtzworth-Monroe, A., (1) 223, 232, 291, 322

Holyoak, K. T., (1) 9, 34, 111, 178, 199, 236, 242, 277, 283. (2) 22, 24, 28, 60, 302, 332, 427, 433, 4 5 7 Homa, D., (1) 163, 199, 243, 279. (2) 23, 24, 62 Homer, P. M ., (2) 414, 457 Hong, S. T ., (2) 430, 432, 433, 457 Hood, J. E., (2) 297, 336 Hoover, C. W ., (1) 76, 93 Hope, D. A., (2) 301, 332 Hordin, T. S., (1) 372, 388, 401, 409 Horowitz, L., (2) 195, 231, 291, 32 7 Horowitz, M. J., (1) 170, 171, 207. (2) 291, 332

Hough, L., (2) 340, 388 House, P., (2) 306, 335 House, R. J., (2) 340, 391 Houston, D. A., (2) 95, 104, 128 Houston, M. J., (2) 407, 408, 454, 457 Hovland, C. I., (2) 71, 79, 82, 88, 113, 114, 128, 130, 136, 138, 176, 185 Howard, D. J., (2) 90, 93, 122, 128, 430, 463

Howard, J., (1) 74, 96, 214, 216, 260, 281. (2) 5, 22, 24, 65 Howard-Pitney, B., (2) 81, 128, 405, 453, 457

Howell, A ., (1) 370, 400, 409 Hoyer, W. D., (2) 401, 457 Hoyt, S. K., (1) 176, 195

482

AUTHOR INDEX

Hryniuk, W. M ., (2) 244, 284 Hu, L ., (2) 25, 64 Hubbard, C ., (2) 167, 179, 185 Huber, J., (2) 401, 402, 412, 424, 430, 447,

Izard, C. E ., (1) 329, 330, 390, 410

J

457, 463

Hubert, S., (1) 56, 92. (2) 85, 120, 420, 451 Huesman, L. R., (2) 87, 128 Huffcut, A. I., (2) 383, 384, 3 9 7 Huh, E ., (1) 266, 267, 277 Hui, C., (2) 410, 455 Hulin, C. L ., (2) 346, 391 Hull, C. L., (1) 158, 199 Hull, J. G., (1) 2, 35. (2) 291, 305, 322, 332 Humphrey, L. L ., (2) 398, 410 Hunt, E., (1) 107, 147 Hunt, J. R., (2) 269, 281 Hunter, J. E ., (2) 358, 395 Huston, T. L., (2) 205, 206, 235 Hurwitz, J., (2) 163, 182, 185 Hutchinson, H. W ., (1) 24, 26, 27, 3 6 Hutchinson, J. W ., (2) 93, 120, 376, 379, 386, 401, 402, 408, 412, 415-417, 420, 430, 451, 460 Huttenlocher, J., (1) 191, 199 Hutton, D. G., (2) 91, 128 Hymes, L., (1) 19, 20, 31, 32

I Ickes, W ., (1) 76, 95. (2) 86, 128 Iffaldano, M ., (2) 340, 391 Ilgen, D. R., (2) 340-342, 347, 350, 351, 362, 365, 366, 369, 391 Ingerman, C., (2) 214, 233 Ingram, R. E., (1) 321, 409. (2) 290, 296-300, 301, 332, 333 Inks, L ., (2) 377, 392 Inn, A., (2) 362, 366 Innis, D. E ., (2) 99, 133, 413, 460 Insko, C. A ., (2) 2, 6, 42, 66, 11, 81, 86, 97, 124, 128 Isen, A. M ., (1) 5, 35, 374, 376, 377, 379-381, 396, 397, 400, 409, 410. (2) 62, 101, 123, 128, 175, 182, 217-219, 232, 235 Isenberg, D. J., (2) 340, 391 Isham, J. T ., (1) 54, 9 7 Iyengar, S., (2) 74, 128, 144, 150, 153, 154, 156-158, 164, 173, 180, 182 Iyer, E. S., (2) 401, 461

Jaccard, J. J., (2) 116, 117, 124, 128, 427, 45 7

Jackson, D. D., (1) 72, 97 Jackson, L. A ., (2) 46, 62 Jacobsen, E ., (1) 381, 410 Jacobson, N. S., (2) 223, 232, 235, 291, 332 Jacoby, L. L., (1) 14, 15, 30, 35, 121, 131, 135, 139, 141, 143, 147. (2) 41, 62, 422, 440, 455 James, L. R., (2) 343, 391 James, W ., (1) 154, 155, 193, 199, 323, 324, 328, 331, 369, 410 Jamieson, D. W ., (2) 104, 128 Janigan, A. S., (2) 291, 3 3 7 Janis, I. L., (2) 82, 87, 128, 129, 244, 245, 283

Janiszewski, C ., (2) 403, 412, 451, 4 5 7 Janoff-Bulman, R., (2) 291, 332 Jasechko, J., (1) 15, 35, 141-143, 147. (2) 41, 62 Jemmott, J. B., (2) 248, 249, 269, 281-283 Jenkins, R., (2) 406, 465 Jensen, C., (1) 74, 96, 214, 216, 260, 281. (2) 5, 22, 24, 65 Jepson, C., (2) 102, 129 Jerdee, T. H ., (2) 368, 394, 395 Jobe, J. B., (2) 261, 262, 285 John, O. P., (1) 5, 16, 17, 19, 20, 38, 160, 799, 214, 215 , 282. (2) 21, 53-55, 61, 65

Johnson, B. T ., (2) 88, 90, 91, 94, 129, 344, 376, 391, 413, 4 5 7 Johnson, C., (2) 12-14, 62, 64 Johnson, D., (1) 159, 162, 205. (2) 103, 122 Johnson, D. J., (2) 203, 204, 227, 228, 235, 236

Johnson, E. J., (1) 383, 410. (2) 102, 129, 174, 755, 401, 402, 412, 418, 420, 430, 445, 452, 457, 461 Johnson, H. H ., (2) 113, 729 Johnson, J., (1) 188, 208. (2) 8 7 , 138, 423, 463 Johnson, J. E., (2) 243-246, 275, 277, 283, 284

Johnson, M ., (1) 397, 410 Johnson, M. K., (1) 2, 35, 86, 95, 298, 319. (2) 325, 332

AUTHOR INDEX

Johnson, R. C ., (1) 385, 397, 417 Johnson, R. D., (2) 424, 430, 45 7 Johnson-Laird, P. N., (1) 273, 27 9 , 327, 329, 336, 363, 412 Johnston, L., (2) 51, 61, 62 Jolly, J., (2) 380, 391 Jones, C. R., (1) 2, 5, 8, 35, 53, 60, 95, 111, 147, 219, 228, 256, 260, 279, 300, 318. (2) 41, 62, 65, 104, 105, 129, 363, 391 Jones, R. A., (1) 168, 199. (2) 15, 62 Joong-nam, Y., (1) 309, 318 Joreskog, K. G., (1) 288, 318 Jordan, J. S., (2) 290, 305, 318, 332, 338 Jose, P. E ., (1) 336, 343-348, 365, 413 Judd, C. M ., (1) 74, 95, 96, 137, 142, 147, 244, 257, 279, 282. (2) 22, 25, 27-31, 62, 64, 77, 79-81, 88, 129, 164, 178, 182, 341, 395, 416, 457 Jussim, L. J., (1) 9, 11, 36. (2) 109, 130, 343, 348, 349, 352, 354, 356-358, 361, 365, 379, 381, 385, 391 Just, M. A., (1) 105, 151

K Kagan, J., (1) 172, 199, 364, 410 Kahan, T. L., (2) 325, 332 Kahle, L. R., (2) 414, 457 Kahneman, D., (1) 3, 15, 35, 36, 115, 147, 191, 208, 214, 243, 264, 266, 284, 303, 304, 306-308, 313, 318, 321, 334, 385, 397, 410, 417. (2) 12, 20, 21, 23, 24, 62, 67, 152, 164, 182, 250, 283, 361, 396, 402-404, 406, 408-410, 420-423, 426, 434, 445-447, 457, 458, 465 Kallgren, C. A., (2) 93, 106, 116, 129, 141, 341, 388 Kane, J. S., (2) 385, 386, 391 Kane, K. F., (2) 385, 386, 391 Kane, M ., (2) 290, 333 Kanfer, R., (2) 340-342, 391 Kanner, A. D., (1) 363, 410. (2) 263, 283 Kant, I., (1) 158, 199 Kao, C. F., (1) 255, 275. (2) 116, 123 Kapiro, J., (2) 267, 283 Kaplan, K. J., (2) 70, 129 Kaplan, M. F., (1) 218, 279 Kaplan, N., (2) 194, 234

483

Kaplow, S., (2) 365, 390 Karasawa, K., (1) 357, 411. (2) 27, 62 Karaza, J., (2) 290, 335 Kardes, F. R., (1) 4, 5, 12, 18, 20, 33. (2) 70, 93, 112, 117, 125, 129, 137, 400, 402, 407, 409, 412, 414, 416, 417, 420, 422-431, 433, 434, 440-442, 448, 455, 456, 458, 462, 464, 465 Kardinal, C. G., (2) 270, 285 Karlin, M. B., (2) 7, 21, 58 Karlins, M ., (2) 49, 62 Kamiol, R., (1) 232, 262, 263, 279 Karp, L ., (1) 5, 35, 314, 376, 380, 397, 410. (2) 175, 182, 218 , 235 Kashima, E. S., (2) 62 Kashima, Y., (2) 27, 62 Kasma, J. A., (2) 99, 107, 108, 134, 135 Kassin, S., (2) 92, 129 Katz, D., (1) 128, 147. (2) 49, 62, 72, 129, 413, 414, 458 Katz, I., (2) 42, 62 Katz, L. B., (1) 51, 52, 55, 56, 61, 63, 94, 180, 198, 260, 278. (2) 302, 331 Katz, P. A., (2) 40, 62 Kauff, D. M ., (1) 9, 38 Kaufman, C. M ., (1) 154, 195 Kavanagh, M ., (2) 383, 384, 390 Kayne, N. T ., (2) 290, 306, 307, 329, 332 Keane, T. M ., (2) 301, 333 Keating, C. F., (1) 186, 199 Keating, J. P., (2) 100, 101, 130 Keech, W ., (2) 149, 180 Keele, S. W ., (2) 16, 22, 24, 64 Keenan, J. M ., (1) 176, 181, 200. (2) 313, 332, 408, 458 Kegan, R., (1) 172, 200 Kehoe, J. F., (2) 379, 388 Keil, F. C ., (1) 158, 200, 237, 247, 279 Keil, L. J., (2) 271, 284 Keith, P. M ., (2) 197, 236 Keller, K. L., (2) 418, 419, 431, 451, 458 Keller, M. L., (2) 251, 253, 284 Kelley, C. M ., (1) 15, 30, 35, 121, 131, 139, 141-143, 147. (2) 41, 62 Kelley, H. H., (1) 87, 95, 161, 200, 382, 410. (2) 82, 104, 128, 129, 202, 203, 205, 206, 218, 223, 228, 230, 235, 238, 302, 332, 350, 391 Kelley, S., (2) 172, 173, 182 Kelly, G. A., (1) 171, 200. (2) 380, 391 Kelly, K. A., (2) 294-296, 320, 32 7

484

AUTHOR INDEX

Kelman, H. C ., (2) 89, 113, 114, 129, 130 Kelsey, R. M ., (2) 72, 121 Keltner, D., (1) 382, 384, 385, 410 Kemeny, M. E ., (2) 228, 23 7 , 264, 2 8 7 Kendall, P. C ., (1) 168, 200. (2) 290, 292, 297, 298, 332, 333 Kennedy, C ., (2) 211, 235 Kennedy, S., (1) 173, 200 Kenny, D. A., (1) 76, 93. (2) 309, 333, 361, 392

Kenrick, D. T ., (2) 103, 140, 219, 232 Kent, G., (2) 261, 283 Kent, R. J., (2) 419, 458 Kerkar, S., (2) 379, 394 Kerker, R. M ., (1) 71, 98. (2) 79, 140 Kern, D ., (1) 310, 321 Kenski, H. C., (2) 115, 135 Kemell, S., (2) 149, 182 Kerr, T ., (2) 142, 145 Kesnaiemi, A., (2) 267, 283 Key, V. O., (2) 147, 182 Kiel, G. C., (2) 401, 458 Kieras, D. E., (1) 107, 112, 147 Kiesler, C. A., (2) 73, 81, 95, 100, 108, 130, 141

Kiewiet, R. D., (2) 147, 148, 152, 161, 182, 183

Kihlstrom, J. F., (1) 42, 132, 134, 137, 138, 145, 154, 156, 157, 162-164, 171, 176, 177, 180, 185, 190, 193, 196, 197, 200, 204, 20 6

Kim, H., (1) 255, 279 Kim, H-S., (2) 44, 62 Kim, J., (2) 407, 414, 427, 430, 440, 441, 456, 458

Kim, J. I., (1) 29, 39 Kimchi, R., (1) 101, 149 Kinder, D. R ., (1) 137, 146. (2) 71, 74, 120, 128 , 147, 148, 150*, 156-158, 160, 161, 170, 173, 176, 179, 181-183 King, B. T ., (2) 87, 129 King, G. A., (1) 6, 11, 16, 35, 44, 53, 95, 163, 199, 219, 228, 246, 279. (2) 40, 61, 154, 161, 166, 167, 182, 266, 28 3 , 302, 305, 331 King, L. A., (1) 335, 408 Kintsch, W ., (1) 76, 9 6 Kirker, W. S., (1) 176, 199, 205 Kirsch, M. P., (2) 360, 381, 392 Kirschenbaum, D. S., (1) 398, 410 Kirscht, J. P., (1) 117, 136

Kirsner, K., (1) 121, 148 Kirson, D ., (1) 335, 336, 338, 358, 359, 414. (2) 201, 2 3 7

Kirtsch, I., (2) 297, 336 Kisielius, J., (2) 84, 130, 407, 434, 435, 458 Kitayama, S., (1) 174, 175, 202 Kite, M. E ., (1) 246, 276 Klapp, S. T ., (1) 106, 148 Klar, Y., (2) 410, 458 Klatzky, R. L ., (1) 73, 92, 234, 275, 279. (2) 35, 5 8 Klayman, J., (1) 132, 134, 147. (2) 348, 392, 428, 443, 458 Klein, C., (2) 99, 124 Klein, D. J., (2) 102, 140 Klein, J. J., (2) 340, 391 Klein, R., (1) 169, 199, 360, 409. (2) 87, 128, 297, 331 Klein, S. B., (1) 42, 55, 56, 95, 157, 163, 176, 177, 180-182, 184, 185, 200, 201, 220, 279. (2) 23, 62 Kleinbard, J., (2) 422, 440, 454 Klimoski, R., (2) 377, 392, 396 Klinger, E ., (1) 128, 147. (2) 116, 183, 404, 458

Klinger, M. R., (1) 9, 10, 12, 20, 34, 302, 319

Klonsky, B. G., (2) 346, 361, 363, 389 Klotz, M ., (2) 270, 282 Klumpp, G., (1) 14, 15, 38, 141, 150, 306, 320, 385, 414 Knapp, A., (2) 219, 235 Knauper, B., (1) 295, 306, 307, 320 Knetsch, J. L ., (2) 446, 4 5 7 Knight, J., (1) 188, 204 Knowles, E. S., (1) 154, 201 Knutson, J. F., (2) 361, 3 8 7 Koenig, L. J., (1) 392, 410. (2) 318, 333 Koenig, O., (I) 186, 201 Kokinov, B., (1) 274, 279 Kolers, P. A., (1) 116, 119, 147, 148 Kolodner, J., (1) 192, 201 Kommer, D., (1) 387, 414 Kopelman, M. D ., (1) 185, 201 Koriat, A., (2) 424, 459 Koskenvuo, M ., (2) 267, 283 Kosslyn, S. M ., (1) 186, 189, 201 Kost, K. A., (2) 406, 455 Kotalik, J., (2) 244, 284 Kothandapani, V ., (2) 176, 183 Kozlowski, L. T ., (1) 186, 201

AUTHOR INDEX

Kozlowski, S. W ., (2) 360, 361, 375, 381,

48 5

L

392

Kraft, D., (2) 72, 81, 117, 140, 141 Kraft, R. N., (2) 405, 459 Kraiger, K., (2) 341, 365-367, 390, 392 Krames, L., (2) 321, 330 Krantz, D. H., (1) 308, 31 7 Krantz, D. S., (2) 244, 283 Krantz, S. E., (2) 216, 234 Kraus, S., (2) 30, 62 Krauss, R. M ., (1) 291, 302, 318, 319 Kraut, R. E., (1) 166, 196 Kravitz, D. A., (2) 370, 392 Krejci, M. J., (2) 291, 334 Kritz-Silverstein, D., (2) 117, 128 Krosnick, J. A., (1) 9, 11, 36, 142, 147. (2) 74, 77, 80, 81, 109, 119, 121, 128-130, 154, 157, 164, 167, 173, 178, 182, 183 Krueger, J., (1) 230, 244, 254, 279. (2) 33, 37-39, 62 Kruglanski, A., (2) 93, 102, 103, 130, 309, 333, 351, 352, 358, 376, 382, 392, 413, 414, 421, 424, 433, 444, 459 Krugman, H. E ., (2) 402, 459 Krull, D. S., (1) 3, 26, 27, 34, 44, 60, 64, 94, 126, 146, 221, 253, 268, 277. (2) 79, 126, 421, 455 Kubler, A., (2) 97, 139 Kuhl, J., (1) 389, 399, 400, 410 Kuhn, M. H., (1) 168, 201 Kuiper, N. A., (1) 175-178, 188, 201, 205, 379, 407. (2) 297, 300, 304, 306, 311, 312, 329, 333, 33 7 Kuipers, P., (1) 353, 408 Kukla, A., (2) 222, 238, 350, 396 Kuklinski, J., (1) 301, 319. (2) 165, 172, 176, 179, 183, 185, 187 Kulick, J. A., (2) 264, 277, 283 Kulik, J. A., (1) 178, 195. (2) 79, 129, 416, 457

Kumar, P. A., (1) 51, 57, 94, 181, 198, 223, 278

Kumpf, M ., (1) 382, 385, 414. (2) 102, 137 Kunda, Z., (1) 135, 148, 167, 170, 202, 221, 230, 279, 400, 410. (2) 33, 38, 63, 94, 130, 274, 283, 343, 351, 368, 372, 376, 377, 382, 385, 392 Kunst-Wilson, W. R., (1) 9, 36. (2) 104, 130 Kussan, M. E., (1) 258, 279 Kuykendall, D., (2) 100, 101, 130

Lacey, B. C., (1) 390, 410 Lacey, J. I. (1) 390, 410 Lachman, J. T ., (2) 160, 183 Lachman, R., (2) 160, 183 Lafay, J., (2) 149, 150, 184 Lage, E., (2) 102, 133 Lahey, M. A., (2) 378, 392 Laird, J. D., (1) 376, 410 Lakatos, I., (2) 424, 459 Lakoff, G., (1) 158, 201 Lamb, M. E., (2) 192, 235 Lambert, A. J., (1) 58, 66-69, 75, 90, 95, 98, 230, 279 Lamm, H., (2) 198, 235 Lamphere, R. A., (1) 154, 201 Lance, C. E., (2) 377, 382 Landy, F. J., (2) 345-347, 368, 385, 388, 392

Lane, D. M ., (2) 369, 393 Lane, J. W ., (1) 371, 408 Lane, R. E., (2) 163, 183 Lang, G., (2) 161, 183 Lang, K., (2) 161, 183 Lang, P. T ., (1) 328, 410 Langer, E. J., (1) 2, 7, 33, 37. (2) 403, 453 Lang-Gunn, L., (2) 291, 332 Langivanio, H., (2) 267, 283 Langston, C., (1) 403, 415. (2) 433, 464 Lansman, M ., (1) 107, 147 Lanza, R. P., (1) 172, 196 Larrick, R. P., (2) 447, 459 Larter, W. M ., (1) 137, 146. (2) 160, 181 Larwill, L. K., (1) 142, 147 Lassiter, G. D ., (1) 392, 410. (2) 318, 333 Lasswell, H. D ., (2) 144, 183 Latham, G. P., (2) 342, 370, 383, 392 Lau, R. L., (2) 149, 152, 153 Lau, R. R., (2) 87, 130, 160, 161, 183, 253, 254, 283 Lau, T., (2) 84, 140 Laude, R., (2) 297, 328 Lauden, L., (2) 424, 459 Laughlin, P. R., (2) 340, 392 Laughunn, D., (2) 152, 185 Lauver, D. R., (2) 244, 283 Lave, L., (2) 251, 260, 284 Lavine, H., (2) 81, 91, 140 Law, J., (1) 119, 150 Lawrence, B., (2) 368, 392

486

AUTHOR INDEX

Lawson, R., (2) 431, 461 Layton, R. A ., (2) 401, 458 Lazarus, R. S., (1) 20, 36, 337, 351, 356, 363, 410. (2) 242, 248, 263, 281, 283, 291, 318, 329, 330 Lazersfeld, P. F., (2) 146, 165, 180, 183 Leary, M. R., (1) 154, 201. (2) 290, 291, 333

Leavitt, C., (2) 91, 127 Lee, J. A., (2) 368, 392 Lee, J. K., (2) 263, 284 Lefcourt, H. M ., (2) 291, 333 Lefebvre, R. C., (2) 264, 2 86 Legant, P., (1) 304, 31 7 Leggett, E. L ., (1) 357, 358, 407 Lehmann, D. R., (2) 401, 430, 452, 468 Lehrer, R., (1) 171, 202 Leight, K. A., (1) 371, 375, 410 Leiker, M ., (2) 267, 283 Leippe, M. R., (2) 91, 96, 113, 114, 116, 125, 130, 136, 167, 183 Leirer, V. O., (1) 51, 52, 55, 56, 61, 63, 94, 180, 198, 260, 278. (2) 302, 331 Lemieux, A. M ., (2) 258, 2 86 Lemley, R. E ., (2) 444, 461 Lemley, R. H ., (1) 402, 412 Lemon, N., (2) 302, 333 Lenauer, M ., (1) 181, 202 Lenn, T. M ., (2) 23, 60 Leone, C., (1) 260, 284. (2) 88, 139 Leone, D. R., (1) 9, 33, 141, 145 Lepper, M. R., (1) 13, 29, 36, 230, 280. (2) 87, 95, 105, 130, 131, 250, 284, 440, 444, 459 Lerma, M ., (2) 227, 23 7 Lemer, M. J., (1) 3, 6, 26, 39, 117, 150, 263, 283. (2) 196-198, 2 33-235 Levenson, R. W ., (1) 329-331, 407, 411. (2) 229, 234 Leventhal, E. A., (2) 251, 253, 285 Leventhal, H ., (1) 179, 202, 330, 411. (2) 243, 245, 246, 249, 251-254, 267, 275, 277, 280, 2 83-285 Levin, I. P., (2) 424, 430, 446, 457, 459 Levin, L ., (2) 244, 284 Levine, J. M ., (2) 103, 130 Levine, M. N., (2) 244, 284 Levine, R. A., (1) 174, 206. (2) 366, 392 Levine, S., (1) 387, 411 Levinger, G., (2) 205, 206, 235 Levinson, D ., (2) 145, 179

Levinson, S. C ., (1) 292, 309-311, 317, 319 Levitt, K., (1) 6, 3 7 Levy, A. S., (1) 2, 35. (2) 427, 454 Levy, S. M ., (2) 263, 284 Lewan, P. C ., (2) 115, 130 Lewandowski, S., (1) 121, 148 Lewicki, P., (1) 2, 8, 36, 107, 147, 239, 273, 279. (2) 19, 20, 24, 63, 302, 333 Lewinsohn, P. M ., (2) 299, 333 Lewis, L. L ., (1) 21, 33. (2) 18, 59 Lewis, M ., (1) 172, 173, 202 Lewis, M. W ., (2) 384, 388 Lewis, S., (1) 214, 230, 245, 282. (2) 54, 65 Lewis, S. K., (2) 437, 454 Lewis-Beck, M ., (2) 149, 150, 183, 184 Li, F., (1) 73, 9 6 Liberman, A., (1) 124, 125, 128, 131, 132, 135, 146, 265, 276. (2) 85, 88, 95, 104, 106, 107, 123, 130, 411, 453 Lichtenstein, M ., (1) 42, 51, 56, 58, 60, 65, 92, 95, 97, 128, 145, 181, 207, 212, 223, 254, 270, 275, 283. (2) 24, 33, 35, 36, 39, 44, 58, 66, 83, 131, 177, 183, 373, 392, 420-423, 431, 436, 440, 442, 443, 453, 459, 464 Lichtenstein, S., (2) 154, 186, 424, 429, 443, 455, 459

Lichtman, R. R., (2) 270, 271, 277, 287, 288 Lim, J., (2) 414, 427, 430, 458 Lin, S., (1) 233, 284 Lindell, M. K., (2) 340, 343, 348, 379, 389 Lindsay, D. S., (1) 14, 35, 298, 319 Lingle, J. H ., (1) 160, 163, 181, 202, 204. (2) 23, 63, 70, 131, 167, 183, 430, 459

Linville, P. W ., (1) 47, 93, 134, 138, 148, 167, 168, 202, 257, 279. (2) 18, 21, 26-28, 63, 81, 131, 250, 257, 264, 274, 278, 284, 291, 333 Lipitz, M. E ., (2) 220, 234 Lipkus, I., (2) 203, 204, 236 Lippman, M ., (2) 263, 284 Lippmann, W ., (2) 4, 25, 42, 63 Lipson, A., (1) 266, 267, 277 Lisle, D. J., (2) 72, 140 Lisman, S. A., (1) 375, 413 Litt, J., (1) 17, 19, 32 Litz, B. T ., (2) 301, 333 Liu, J. H., (1) 357, 411 Liu, T. J., (1) 9, 10, 12, 20, 34 Lloyd, B. B., (1) 157, 205

AUTHOR INDEX Lloyd, S. A., (2) 197, 232, 235 Lobel, M ., (2) 271-273, 277, 28 7 Locke, E. A., 340, 342, 392 Locke, K. D., (1) 382, 384, 410 Lockhart, M ., (2) 362, 370, 393, 394 Lockhart, R. S., (1) 54, 66, 93, 176, 196 Locksley, A., (1) 181, 202, 253, 254, 279. (2) 37-39, 63, 361, 392 Lodge, M ., (2) 166, 167, 169, 170, 172, 173, 177, 179, 183, 184 Loebel, M ., (2) 214, 2 3 7 Loftus, E. F., (1) 44, 60, 86, 93, 95, 192, 202, 298, 302, 319. (2) 168, 181, 256, 261, 281, 418, 454 Loftus, J. B., (1) 55, 56, 95, 163, 176, 177, 180-182, 184, 185, 201, 220, 279. (2) 23, 62 Logan, G. D., (1) 3, 16, 31, 36, 105, 106, 108, 114, 115, 117-119, 122, 137, 148, 151, 179, 202, 263, 272, 280. (2) 308, 333 Logan, H., (1) 255, 275 Loken, B. A., (1) 42, 95 Lombardi, W. J., (1) 5, 8, 10-12, 32, 36, 44, 53, 54, 92, 94, 95, 256, 278. (2) 8, 58, 61, 166, 180, 416, 456 Longnecker, C. O., (2) 342, 345, 362, 372, 377, 392 Longo, L. C ., (2) 367, 389 Lopatka, C., (2) 261, 275, 282 Lopes, L., (1) 217, 280. (2) 340, 393 Lopez, A., (1) 223, 240, 241, 2 80 Lopez, D. F., (1) 9, 12, 32 Lord, C. G., (1) 13, 29, 36, 230, 280. (2) 10, 63, 95, 131, 250, 284, 440, 444, 459

Lord, F. M ., (1) 289, 319 Lord, R. G., (2) 340, 342, 354, 359, 373, 375, 381, 390, 393, 394 Losch, M. E., (2) 96, 131 Louie, D., (2) 248, 281 Love, R. E., (2) 86, 131 Lucas, J., (2) 365, 390 Lucia-Irizarry, N ., (1) 174, 198 Lui, L. N., (1) 23, 33, 251, 276. (2) 7, 17, 18, 50-52, 59 Lujanski, H., (2) 197, 235 Lumsdaine, A. A., (2) 113, 128 Lunn, R. O., (2) 368, 395 Lupfer, M. B., (1) 24, 26, 27, 3 7 Lurie, L., (1) 313, 318

4 87

Luschene, R. E ., (2) 313, 320, 336 Lusk, C. M ., (2) 88, 129 Lutz, C., (1) 330, 353, 411 Lyden, J., (2) 113, 131 Lynch, J. G., (2) 83, 126, 343, 349, 351, 356, 368, 376, 379, 385, 389, 415, 425, 427, 438-441, 451, 455, 459, 463

Lynch, L., (1) 227, 269, 283 Lynch, M ., (2) 96, 139 Lynn, A. R., (1) 9, 11, 36. (2) 109, 130 Lynn, S., (1) 312, 320 Lyon, D ., (2) 291, 333 Lyons, P. M ., (2) 368, 388

M Maass, A., (2) 6, 13, 63, 65, 102, 103, 131, 140

MacCarthy, B., (2) 291, 328 MacDonald, C., (2) 304, 328 MacDonald, M. R., (2) 297, 300, 304, 306, 311, 312, 333 MacDougall, J. M ., (2) 266, 267, 277, 281, 285

Mackenzie, S. B., (2) 78, 84, 99, 131, 345, 358, 393, 407, 459 Mackie, D. M ., (1) 14, 16, 38, 212, 219, 230, 255 , 280, 282, 283, 372, 389, 393 , 405, 411, 416. (2) 6, 9, 14, 19, 22, 23, 27-30, 43^15, 56, 61, 63, 64, 67, 73, 84, 86, 90, 100-103, 108, 109, 113, 130, 131, 138, 141, 411, 431, 433, 460 MacLeod, C., (2) 292, 293, 296, 298, 299, 301, 302, 330, 333, 338 Macrae, C. N ., (2) 24, 34-36, 39, 40, 42, 63, 64

Madden, T. J., (2) 412, 422, 451, 459 Maddux, J. E., (2) 290, 291, 334, 33 7 Madigan, R. J., (1) 376, 397, 411, 412 Madigan, S. A., (2) 426, 459 Magovem, G. J., (2) 264, 286 Maguire, E ., (2) 362, 394 Mah, W. A., (1) 177, 199 Maher, K. J., (2) 340, 393 Maheswaran, D., (2) 92, 95, 123, 132, 406, 411, 431, 433, 460 Mahler, H. I. M ., (2) 264, 283 Mahoney, M. J., (2) 290, 334

488

AUTHOR INDEX

Mai, H. P., (2) 300, 320 Maier, S. F., (1) 357, 414 Main, M ., (2) 194, 234 Major, B., (2) 47, 59, 64 Makhijani, M. G ., (2) 346, 361, 363, 367, 389

Malanchuk, O., (2) 170, 184 Malcolm, T ., (2) 96, 126 Malone, P. S., (2) 79, 126 Malpass, R. S., (2) 84, 140 Malt, B. C., (1) 140, 148. (2) 23, 64 Mandler, G., (1) 9, 19, 36, 141, 148, 336, 337, 363, 411. (2) 92, 132, 403, 406, 460

Mandler, J., (1) 47, 77, 95, 191, 202, 371, 411

Manis, M ., (1) 215, 224, 280, 310, 319. (2) 33, 39, 58, 168, 184, 436, 463 Manning, D. J., (1) 309, 311, 3 1 7 Manstead, A. S. R., (1) 337, 350, 411, 412 Maoz, Y., (1) 27, 40 Maracek, J., (1) 304, 319 Marato, J. J ., (2) 264, 284 Marburger, W ., (1) 192, 202 Marcel, A. J., (1) 17, 3 6 Marchese, L. A., (1) 171, 200 Marco, C. A ., (1) 154, 207. (2) 270, 271, 286

Marcus, G. E., (2) 163, 184, 186 Martell, R. F., (2) 364, 369, 381, 390, 393 Mark, M. M ., (1) 5, 36, 394, 396, 415. (2) 101, 138 Markman, E. M ., (1) 158, 202 Marks, L. E., (1) 272, 280 Marks, L. J., (2) 93, 136 Marks, T ., (2) 300, 331 Markus, G. B., (1) 7, 40. (2) 149, 184 Markus, H., (1) 2, 36, 154, 156, 165-167, 170, 174, 175, 196, 197, 202, 219, 246, 277, 280. (2) 72, 73, 109, 132, 142, 293 , 302, 303 , 308, 309, 324, 329, 330, 334, 335, 342, 365, 379, 389, 393

Marlatt, G. A., (2) 291, 33 7 Marlino, D., (2) 407, 465 Marmorstein, H., (2) 93, 106, 120, 439, 440, 459

Marrs, S., (1) 223, 283. (2) 35, 66 Marsh, K. L., (1) 391, 392, 395, 411, 416. (2) 295, 317-322, 334, 338 Marsh, R. L ., (1) 9, 38

Marshall-Garcia, K. A., (1) 376, 411 Marshall-Goodell, B. S., (2) 108, 122 Martin, C ., (2) 379, 394 Martin, D. J., (2) 311, 334 Martin, E., (1) 184, 202, 289, 316, 319, 321 Martin, L. L ., (1) 12, 28, 36, 54, 57-60, 64, 71, 95, 97, 98, 127, 131, 148, 290, 299, 315, 316, 321, 385, 390, 391, 397, 399, 400, 402, 411, 415. (2) 87, 97, 108, 132, 138, 168, 169, 184, 187, 240, 278, 384 Martin, N. G., (2) 70, 125 Martindale, C ., (1) 161, 203 Martz, J., (2) 203, 204, 236 Marwell, G., (2) 70, 132 Maser, J. D ., (2) 291, 334 Maruyama, G., (2) 346, 361, 362, 396 Marz, P., (1) 395, 405 Masters, J. C., (1) 398, 399, 411. (2) 271, 284

Mathews, A., (2) 376, 411. (2) 292, 293, 296-299, 301, 302, 314, 328, 330, 333, 334, 338

Mathur, M ., (1) 101, 132 Matlin, M. W ., (1) 379, 411. (2) 110, 132 Matter, J., (1) 392, 406 Matthews, K., (2) 264, 265, 284, 286 Maurer, T. J., (2) 346, 364, 370, 393 Mauro, R., (1) 353, 411 Mavin, G. H ., (1) 16, 35, 219, 228, 246, 279. (2) 302, 305, 331 May, J., (2) 302, 334 May, M. A., (1) 166, 198 Mayer, J. D., (1) 276, 380, 383, 406, 411. (2) 102, 132 Mayol, R., (2) 300, 331 Mazur, A., (1) 186, 199 Mazursky, D., (2) 114, 132, 436, 460, 463 Me Andrews, M. P., (1) 119, 148 McArthur, L. Z ., (1) 12, 36, 186, 203, 218, 275, 280

McCabe, S. B., (2) 298, 331 McCallum, M. C ., (2) 418, 422, 465 McCann, C. D., (1) 66, 95, 161, 195, 292, 311, 318, 319. (2) 94, 128, 132 McCann, J. W ., (2) 424, 430, 457 McCaul, K. D., (1) 395, 411. (2) 318, 334 McClelland, D. C ., (2) 145, 184 McClelland, J. L ., (1) 110, 111, 118, 135, 146, 148, 149

McClintock, E ., (2) 205, 206, 235

AUTHOR INDEX McCloskey, M ., (1) 298, 319 McCormick, L. A., (2) 321, 322, 338 McDonald, D. W ., (2) 223, 235 McDonel, E. C., (2) 448, 454 McDougall, W ., (1) 329, 411 McEvoy, C. L., (2) 423, 431, 461 McEachem, L ., (2) 292, 317, 338 McFarland, A. D ., (1) 371, 408 McFarland, C., (2) 115, 137 McGill, A., (2) 244, 282, 407, 460 McGill, G. A., (1) 304, 305, 319 McGinnies, E ., (2) 113, 133 McGraw, K., (2) 163, 165-167, 169, 170, 172, 173, 177-179, 184, 214, 233 McGuire, C. V., (1) 42, 44, 91, 95, 169, 170, 172, 173, 203. (2) 78, 87, 90, 132, 161, 162, 184 McGuire, W. J., (1) 7, 36, 42-44, 91, 91, 95, 169, 170, 172, 173, 203. (2) 75, 77, 78, 80-83, 87, 89, 90, 112, 114, 115, 132, 133, 144, 161, 162, 182, 184

McHoskey, J. W ., (2) 95, 133 Mclntire, M. D., (2) 380, 393 McIntosh, W. D., (2) 213, 238 McKoon, G., (1) 214, 280, 418, 462 McKuen, M. B., (2) 147, 156, 184 McMahon, S., (2) 249, 253, 255, 280 McManus, J., (2) 256, 285 McMillan, D., (2) 34, 35, 37, 66 McMullen, M. N., (2) 400, 463 McNally, R. J., (2) 297, 332 McNamara, T. P., (1) 160, 240, 418, 460 McNeil, B., (2) 152, 184 McNulty, S. E ., (2) 291, 3 3 7 McPartland, T. S., (1) 168, 201 McPhee, W. N., (2) 146, 180 McQuinn, R. D ., (2) 197, 219, 231 McSweeney, F. K., (2) 108, 133, 412, 453, 460

Means, B., (1) 391, 410 Mebane, W. R., (2) 150, 183 Mecklenbrauker, S., (1) 380, 411 Medin, D. L., (1) 139, 140, 148, 157-160, 162-164, 2 0 2-204, 206, 226, 243, 259, 270, 272, 280. (2) 20, 21, 23, 64, 254, 286, 343, 385, 393 Meehl, P. E., (1) 232, 276 Mees, U., (1) 324, 328, 411 Meglino, B. M ., (2) 341, 347, 350, 351, 372, 388, 393

489

Mehta, P., (2) 199, 235 Melamed, B. G., (2) 263, 286 Melara, R. D ., (1) 137, 148 Melton, R. J., (2) 172, 179, 187 Meltzer, S. J., (2) 309, 331 Melzack, R., (1) 187, 203. (2) 247, 284 Mendel, R. M ., (2) 380, 391 Mendelson, M ., (1) 371, 407. (2) 313, 328 Mendolia, M ., (2) 87, 139 Meneses, L. M ., (2) 44, 63 Merluzzi, T. V., (1) 137, 149. (2) 291, 334 Mertz, E ., (1) 397, 410 Mervis, C. B., (1) 157, 159, 162, 203, 205. (2) 416, 462 Messick, D. M ., (1) 212, 280. (2) 6, 9, 56, 58, 64

Metalsky, G. I., (2) 293, 294, 301, 317, 326, 334

Metcalf, B. L., (2) 400, 462 Meyer, D., (2) 251, 284 Meyer, D. E., (1) 163, 182, 203 Meyer, J. D., (2) 262, 280 Meyer, R. J., (2) 425, 430, 443, 460 Meyerowitz, R. E ., (2) 244, 250, 284 Meyers, A. W ., (1) 399, 412 Meyers-Levy, J., (2) 92, 133, 405, 406, 431, 460

Michaela, J. L ., (1) 161, 200. (2) 274, 284 Middlestadt, S. E ., (2) 167, 185 Mikow, V., (2) 270, 282 Mikula, G., (2) 197, 235 Mikulincer, M ., (2) 194, 235 Milberg, S. J., (2) 39, 60. (2) 199, 217, 232, 431, 461 Millar, K. U ., (2) 74, 133 Millar, M. G., (1) 223, 283. (2) 35, 66, 71, 72, 74, 88, 117, 133, 176, 184, 213, 238 Miller, A., (2) 156, 181 Miller, A. G., (1) 218, 280. (2) 95, 133 Miller, A. H., (2) 170, 184 Miller, A. S., (2) 400, 453 Miller, D. T ., (1) 304, 313, 318. (2)20, 21, 23, 24, 62, 196, 198, 235, 310, 334, 406, 423, 447, 45 7 Miller, F. D., (1) 2, 39, 135, 150 Miller, G. A., (1) 295, 319. (2) 400, 404, 460

Miller, H., (1) 375, 416 Miller, L ., (1) 84, 96 Miller, N., (1) 257, 280. (2) 4, 9, 46, 64, 67, 73, 81, 127, 130

490

AUTHOR INDEX

Miller, R. S., (2) 290, 296, 333 Miller, S. M ., (2) 245, 28 4 , 297, 317, 320, 330

Miller, W. E ., (2) 148, 165, 180 Miller, W. R., (2) 317, 331 Millon, C ., (2) 266, 280 Mills, J., (2) 196, 198-200, 232, 235 Milne, A. B., (2) 24, 34, 39, 40, 42, 63, 64 Milojkovic, J. D., (1) 233, 276 Mineka, S., (2) 292, 294-296, 301, 320, 327, 334

Miniard, P. W ., (2) 91, 99, 100, 133, 407, 413, 460 Minsk, E ., (1) 219, 283 Mirels, H. L ., (2) 290, 338 Mirer, T ., (2) 172, 173, 182 Mischel, T ., (1) 154, 203 Mischel, W ., (1) 160, 196, 374, 412. (2) 17, 59, 293, 329 Mishal, M ., (1) 218, 280 Mita, T. H., (1) 188, 204 Mitchell, A. A., (2) 407, 408, 448, 452, 460 Mitchell, D. B., (1) 119, 148 Mitchell, J. E., (1) 397, 412 Mitchell, M ., (1) 237, 280 Mitchell, M. L ., (1) 24, 38 Mitchell, T. R., (2) 347, 393 Mitnick, L., (2) 113, 133 Miyake, K., (2) 272, 288 Mizerski, R. W ., (2) 441, 460 Mobley, W. H., (2) 341, 393 Mock, J., (2) 313, 328 Modigliani, V., (2) 422, 440, 460 Mogg, K., (2) 301, 302, 334 Monaco, G. E., (2) 409, 410, 456 Mongillo, M ., (2) 361, 392 Monroe, S. M ., (2) 291, 334 Monson, T. C., (1) 161, 204 Montano, D. E ., (2) 81, 116, 124 Monteiro, K. P., (1) 375, 376, 406. (2) 174,

Morey, L. C., (1) 160, 204 Morgan, J. N., (2) 447, 459 Morgan, M. G., (2) 251, 260, 284 Morgan, R. D ., (2) 266, 280 Morganstem, O., (2) 445, 465 Morris, C. D., (1) 121, 122, 148 Morris, K. J., (2) 84, 91, 94, 123 Morris, W. N ., (1) 327, 370, 376, 377, 380, 412

Morrow, G. D ., (2) 203, 236 Moscovici, S., (2) 102, 103, 173 Moscovitch, M ., (1) 119, 148 Moylan, S. J., (2) 44, 60, 111, 234 Mucha, Z ., (2) 416, 465 Muchinsky, P., (2) 340, 391 Mueller, D. R., (1) 168, 197 Mugny, G., (2) 77, 102, 133 Mullen, B., (2) 12, 14, 25, 62, 64, 244, 285 Muller, G., (1) 294, 320 Mulvey, D., (1) 173, 198 Munch, J. M ., (2) 93, 133, 139, 402, 412, 461

Murphy, D. L ., (1) 375, 416 Murphy, G. L ., (1) 162, 163, 204. (2) 23, 64, 343, 385, 393 Murphy, K. R., (2) 342, 344, 346, 347, 354-357, 359, 362, 386, 388, 393, 394

Murphy, S. T ., (1) 9, 11, 37, 385, 412. (2) 109, 133 Murphy, T. D., (1) 139, 140, 148, 243, 280. (2) 23, 64 Murray, N., (1) 396, 397, 412. (2) 101, 133 Murray, S. L., (2) 226, 236 Musante, L., (2) 267, 277, 285 Musson, R. F., (2) 299, 334 Myers, D., (2) 346, 361, 362, 39 6 Myers, G. E., (1) 154, 155, 204 Mynatt, C. R., (2) 444, 454

180

Monteith, M. J., (1) 21, 33. (2) 46, 59 Moore, D. J., (2) 95, 123 Moore, D. L., (2) 91, 99, 100, 133, 401, 402, 412, 460 Moore, J., (2) 213, 238, 111, 2 8 7 Moore, J. C., (1) 310, 319 Mordkoff, J. T ., (1) 162, 204 Moreland, R. L ., (1) 12, 36, 116, 202 Moretti, M. M ., (1) 114, 148. (2) 300, 310, 334

N Naderer, G., (1) 306, 320 Naffrechoux, M ., (2) 102, 133 Nail, L. M ., (2) 244, 283 Naime, J. S., (1) 137, 148 Nakamoto, K., (2) 416, 453 Nakamura, G. V., (1) 47, 84, 92, 94, 179, 191, 195, 204. (2) 246, 280, 293, 328, 348, 38 7

AUTHOR INDEX

Nakamura, Y., (1) 9, 19, 36, 141 ,fc148. (2) 403, 460 Namir, S., (2) 244, 248, 285 Natale, M ., (1) 376, 377, 412 Nathan, B. R., (2) 346, 354, 358, 359, 379, 380, 394 Neale, M. A., (2) 340, 394 Nebergall, R. E., (2) 79, 138 Neccarato, M. E ., (2) 24, 39, 58 Nedler, B. F., (2) 44, 63 Nedungadi, P., (2) 400, 405, 416, 418, 460 Neeley, J. H ., (1) 18, 37, 182, 205 Neisser, U., (1) 4, 18, 37, 155-157, 178, 189, 193, 204, 208. (2) 342, 394 Nelson, C., (1) 230, 277. (2) 364, 390 Nelson, D. L ., (2) 423, 431, 461 Nelson, G., (1) 136, 144. (2) 223-225, 234 Nelson, T. F., (2) 33, 39, 58, 168, 184 Nelson, T. O., (1) 179, 204. (2) 426, 461 Nerenz, D., (2) 251, 267, 284 Neuberg, S. L ., (1) 8, 9, 13, 14, 28, 29, 34, 37, 44, 73, 93, 125, 126, 138, 146, 148, 216, 226, 244, 250, 252, 253, 261, 269, 277, 394, 408 Neumann, D., (2) 170, 184 Neve, P., (2) 103, 133 Neves, D. M ., (1) 112, 148 Newcomb, T. N., (1) 42, 43, 91, 246, 280, 310, 319 Newell, A. W ., (1) 107, 110, 113, 114, 117, 118, 148, 149. (2) 404, 461 Newman, J., (1) 219, 280. (2) 429, 461 Newman, L. S., (1) 12, 23, 24, 26, 37, 40, 267, 280, 360, 412. (2) 348, 351, 394 Newtson, D. A., (1) 42, 77, 78, 96, 392, 412. (2) 318, 334, 348, 370, 394 Nezu, A. M ., (2) 290, 334 Nezu, C. M ., (2) 290, 334 Niedenthal, P., (1) 9-11, 21, 37, 157, 162, 168, 200, 204 Niewenhuyse, B., (1) 328, 412 Nisbett, R. N., (1) 2, 11, 13, 15, 35, 37, 60, 96, 161, 199, 218, 233, 245, 278, 280, 303, 304, 318, 319, 402, 403, 412, 415. (2) 105, 130, 164, 173, 184, 402, 406, 408, 410, 427, 428, 433, 444, 447, 457, 459, 461, 464 Nitz, M ., (2) 115, 135 Nix, G., (2) 299, 301, 335 Noe, A. W ., (2) 363, 395 Noelle-Neumann, E., (1) 295, 306, 307, 320

491

Nohre, L ., (2) 23, 24, 62 Nolan, C. A., (2) 400, 426, 462 Noller, P., (2) 194, 229, 233, 236 Norman, R., (2) 81, 116, 133 Norpoth, H ., (2) 149, 150, 184 Norwood, M ., (2) 81, 116, 124 Norvell, K. T ., (2) 261, 262, 285 Nosofsky, R. M ., (1) 139, 140, 149, 163, 204, 243, 269, 272, 280. (2) 20, 22, 64

Novick, M. R., (1) 289, 319 Nowicki, G. P., (1) 391, 397, 409, 410. (2) 101, 128 Nowlis, H. H ., (1) 381, 412 Nowlis, V., (1) 381, 412 Nurius, P., (1) 156, 166, 202 Nystedt, L ., (1) 166, 204

o Oakes, P. J., (2) 4, 64 Oatley, K., (1) 327, 329, 336, 363, 412. (2) 292, 334 Obermiller, C., (2) 110, 133 Obrist, P. A., (1) 390, 412 O’Connor, C., (1) 335, 336, 338, 358, 359, 414. (2) 207, 23 7 O’Connor, R., (2) 384, 386 Odbert, H. S., (1) 159, 194 Oden, G. C., (1) 157, 204 Offenberg, L., (1) 328, 412 Ohta, N., (1) 119, 150 Oldendick, R. W ., (2) 155, 180 O’Leary, A., (2) 262, 275, 285 O’Leary, K. D., (2) 222, 223, 234 Oleson, K. C., (1) 325, 405 Olian, J., (2) 363, 394 Olien, C., (2) 160, 186, 187 Olinger, L. J., (2) 297, 300, 333 Oliver, R. L., (2) 93, 122, 406, 461 Olney, T. J., (2) 70, 125, 407, 461 Olsen, D., (1) 171, 200 Olshan, K., (1) 56, 96 Olshavsky, R. W ., (1) 162, 196. (2) 401, 454, 461

Olson, J., (2) 75, 133 Olson, J. C., (2) 407, 411, 456, 460 Olson, R., (1) 188, 204, 2 0 7 Omoto, A. M ., (2) 81, 128, 206, 231, 405, 45 7

492

AUTHOR INDEX

Oppler, S. H ., (2) 346, 358, 363, 366, 367, 387, 394

Orbach, J., (1) 188, 204, 2 0 7 Ome, M ., (1) 303, 319 Ortiz, V., (1) 253, 279. (2) 361, 392 Ortony, A., (1) 324, 326, 328-330, 332-339, 341, 342, 348, 352, 359, 361-363, 366-369, 404, 412. (2) 253, 285 Osborne, R. E., (1) 24, 26, 34 Osgood, C. E., (2) 70, 94, 134 Osherson, D. M ., (1) 223, 240, 241, 280 Osterhouse, R. A., (2) 86, 134 Ostrom, C. W ., (2) 147, 184 Ostrom, T. M ., (1) 24, 38, 42, 73, 94, 96, 180, 181, 198, 204, 250, 281, 291, 312, 322. (2) 23, 25, 63, 64, 70, 83, 90, 124, 131, 135, 156, 183, 185, 290, 335, 411, 414, 423, 428, 430, 459, 461, 464

O’Sullivan, C. S., (1) 255, 256, 280. (2) 250, 285

Ottati, V. C., (1) 301, 319. (2) 160, 165, 167, 170-172, 174-176, 183-185, 187 Ouellette, R., (2) 199, 232 Outerbridge, A. N., (2) 358, 395 Owens, J. F., (2) 264, 286

P Pace, T ., (2) 290, 3 3 7 Padawer-Singer, A., (1) 170, 203 Page, B. I., (2) 147, 167, 180, 185 Paivio, A., (1) 186, 204 Pallak, S. R., (2) 107, 134 Palmer, J. C., (1) 60, 86, 95 Pampe, H ., (1) 380, 382, 408. (2) 217, 233 Panksepp, J., (1) 329, 330, 412 Papanek, M. L ., (1) 392, 406 Papageorgis, D ., (2) 115, 133 Parducci, A., (1) 297, 319 Park, B., (1) 24, 35, 42, 59, 60, 65, 74, 9 4 -9 6 , 135, 149, 180, 181, 198, 211, 215, 219, 220, 224, 229, 241, 244, 247, 257, 2 78 -2 8 2 , 373, 381, 409. (2) 18, 22, 24, 25, 27-31, 54, 55, 61, 62, 64, 65, 83, 127, 177, 181, 266, 282, 302, 319, 331, 335, 341, 373, 390, 395, 420-423, 436, 456 Park, C. W ., (2) 401, 402, 411, 417, 431, 452, 461

Parker, L. D ., (2) 92, 120 Parker, S., (2) 152, 184 Parkinson, B., (1) 337, 412 Parrott, W. G., (1) 328, 330, 374, 377-379, 381, 385, 406, 407, 412 Pasahow, R. J., (2) 299, 335 Paskewitz, J. R., (1) 215, 224, 280 Patrick, R., (1) 391, 410 Patterson, M. B., (2) 302, 336 Paulhus, D. L., (1) 6, 37, 289, 319. (2) 213, 238

Pavelchak, M. A., (1) 226, 229, 250, 277. (2) 24, 60, 99, 109, 119, 126, 164, 181, 349, 362, 389, 402, 412, 419-432, 455, 461 Paykel, E. S., (2) 292, 338 Payne, J., (2) 401, 402, 412, 463 Payne, J. W ., (2) 152, 174, 185, 417, 418, 445, 447, 452, 457, 461 Payton, D. W ., (1) 133, 149 Pazy, A., (2) 365, 394 Pbert, L. A., (2) 264, 284 Peak, H., (2) 77, 134 Pearlstone, Z ., (1) 6, 40. (2) 434, 464 Pederson, J. H ., (1) 309, 311, 31 7 Peevers, B., (1) 172, 204, 206 Peffley, M. A., (2) 147, 163, 182, 185 Pelham, B. W ., (1) 3, 26, 27, 34, 44, 60, 64, 94, 126, 146, 221, 253, 268, 277. (2) 81, 134, 291, 315, 335, 421, 455 Pence, E. C., (2) 383, 38 7 Pennebaker, J. N., (2) 243, 250, 251, 265, 282, 2 8 5 -2 8 7

Pennington, N., (1) 85, 86, 96, 191, 205, 221, 281. (2) 161, 185, 379, 394 Penrod, S., (2) 102, 139 Peplau, L. A., (2) 205, 206, 235 Peracchio, L. A., (2) 405, 460 Perdue, C. W ., (1) 6, 9, 20-22, 3 7 Perez, J. A., (2) 77, 102, 133 Perkins, S. C., (1) 399, 412 Perloff, L. S., (2) 271, 285, 291, 335 Perloff, R. M ., (2) 86, 134 Perrig, W ., (1) 76, 9 6 Perrin, N. A., (1) 139, 145 Pervin, L. A., (1) 162, 205 Pessemier, E ., (2) 425, 465 Peters, L. H ., (2) 374, 388 Peters, M. D., (2) 71, 74, 120, 128, 173, 176, 179, 182 Peters, S. D., (2) 208, 234

AUTHOR INDEX

Peterson, C ., (2) 265, 285 , 301, 334 Peterson, D ., (1) 395, 408. (2) 319, 330 Peterson, D. K., (2) 429, 461 Peterson, D. R., (2) 205, 206, 235 Peterson, L ., (2) 251, 280 Petrocik, J. R., (2) 154, 185 Petterson, P., (2) 94, 140 Pettigrew, T. F., (2) 34, 37, 53, 64, 65 Petty, R. E ., (1) 29, 33, 44, 54, 93, 96, 100, 124, 149, 254, 265, 281, 393, 413. (2) 70-74, 78, 80, 81, 83-86, 88, 90-95, 97-104, 106-108, 110-113, 115-117, 119, 121-124, 126, 127, 134-137, 140, 156, 177, 185, 344, 376, 382, 394, 402, 405, 411-414, 423, 424, 453, 461-463 Pfau, M ., (2) 115, 135 Pfeffer, J., (2) 342, 388, 395 Pfeifer, R., (1) 336, 369, 413 Pfeiffer, C., (2) 270, 271, 279 Philbin, T ., (2) 374, 394 Phillips, G., (2) 291, 328 Phillips, J. S., (2) 340, 393 Phillips, K. A., (1) 187, 205 Pierson, J. E., (2) 163, 186 Pieters, R. G. M ., (2) 77, 116, 135 Pietromonaco, P., (1) 2, 8, 10, 11, 32, 53, 92. (2) 40, 58, 257, 280, 308-311, 335, 352, 386 Pilkington, C. J., (2) 213, 238 Pilkonis, P. A., (2) 100, 108, 142 Pillemer, D. B., (1) 156, 193, 208 Pinney, N., (2) 163, 165, 169, 170, 178, 184 Pittenger, J. B., (1) 186, 205, 206 Pittman, N. L ., (1) 391, 413 Pittman, T. S., (1) 10, 11, 30, 33, 37, 166, 199, 391, 413. (2) 91, 135 Pitz, G. F., (2) 429, 461 Pleban, R., (2) 213, 236 Pliner, P., (2) 319, 321, 3 30 Plog, A. E ., (1) 184, 201 Plott, C., (2) 152, 181 Podsakoff, P. M ., (2) 345, 358, 393 Poling, J., (2) 291, 338 Poison, M. C ., (2) 403, 455 Pomerantz, E. M ., (2) 72, 84, 123 Pomerantz, J. P., (1) 186, 201 Pontes, M. J., (2) 422, 458 Pope, M. K., (2) 265, 266, 286 Posner, M. I., (1) 2, 7, 18, 37. (2) 16, 22, 24, 64

493

Postman, L., (1) 14, 3 7 Poulton, J. L., (2) 265, 286 Powell, M. C., (1) 4, 5, 12, 18, 20, 33, 142, 149, 264, 277. (2) 70, 112, 725, 199, 232, 405, 448, 449, 455 Pozo, C ., (2) 265, 280 Pratkanis, A. R., (1) 161, 175, 176, 195, 197. (2) 71, 79, 80, 114, 135, 292, 331, 416, 462 Pratto, F., (1) 3-5, 12, 13, 16, 17, 19-22, 24, 25, 27, 31, 32, 37, 38, 125, 145, 215, 254, 281. (2) 71, 121 Preisler, R. M ., (2) 93, 106, 141 Prentice, D. A., (2) 73, 120, 136,302, 313, 323-325, 335 Presson, C. C ., (1) 162, 196 Preston, E ., (1) 13, 29,36. (2) 250,284, 440, 444, 459 Price, H. D ., (2) 149, 180 Priester, J. R., (2) 106-108, 123, 135, 136 Procidano, M. E ., (2) 217, 236 Prohaska, T. R., (2) 251, 253, 285, 321, 338 Prohaska, V ., (1) 191, 199 Pryor, J. B., (1) 24, 38, 137, 149, 181, 204, 213, 250, 281. (2) 13, 65, 256, 257, 285

Puckett, J. M ., (2) 100, 136 Pulakos, E. D., (2) 346, 358, 363, 367, 383, 387, 394, 396

Punj, G. N., (2) 401, 455, 462 Purcell, E. D., (2) 370, 392 Puto, C ., (2) 447, 457 Pylyshyn, Z. W ., (1) 186, 205 Pyszcynski, T. A., (2) 262, 285, 291, 299, 301, 307, 335, 336

Q Quattrone, G. A., (1) 60, 96, 218, 229, 281. (2) 25-27, 29, 62, 65, 152, 153, 185 Quigley, K. L ., (2) 261, 282 Quillian, M. R., (2) 164, 181 Quina-Holland, L ., (1) 193, 196 Quine, W. V. O., (1) 241, 281 Quinlan, D. M ., (2) 292, 304, 328

R Raaijmakers, J. G., (1) 177, 205. (2) 418, 462 Rabinowitz, V. C ., (2) 290, 335

494

AUTHOR INDEX

Rachman, S., (2) 261, 275, 282 Radcliff, B., (2) 149, 185 Radecki, C ., (2) 117, 128 Raden, D., (2) 81, 136 Ragins, B. R., (2) 365, 394 Rahn, W. M ., (2) 170, 185 Rasinski, K. A., (1) 253, 254, 28 1 , 299, 321. (2) 70, 80, 140 , 385, 3 9 6 Ratchford, B. T ., (2) 401, 464 Ratcliff, R., (1) 214, 280. (2) 418, 462 Rathjen, D. P., (2) 296, 330 Ratneshwar, S., (2) 93, 136, 402, 411, 412, 416, 462 Raulin, M. L., (1) 187, 188, 196 Ray, M. L ., (2) 86, 93, 121, 402, 411, 412, 452

Raye, C. L ., (1) 86, 95 Raymond, P., (1) 19, 20, 31, 32 Razran, G. H., (2) 108, 136 Rea, C ., (2) 95, 123 Read, S., (1) 76, 84, 92, 96, 236-239, 241, 243, 247, 281 Read, S. J., (2) 194, 196, 233, 379, 394 Reardon, K., (2) 271, 272, 2 8 7 Reardon, R., (2) 84, 91, 95, 123, 133, 136 Redd, W. H., 263, 2 8 7 Reddy, M ., (2) 92, 129 Reed, L ., (2) 222, 238, 350, 3 9 6 Reed, M. J., (2) 299, 332 Reed, S. K., (1) 102, 139, 140, 149 Reeder, G. D., (1) 212, 219, 276, 281, 395, 408. (2) 55, 65, 256, 257, 285, 302, 319, 330, 335 Regan, D. T ., (2) 81, 136 Rehm, L. P., (2) 297, 298, 335 Reichel, L. D., (2) 44, 63 Reilly, N. P., (1) 377, 412 Reilly, R. R., (2) 346, 355, 375 Reimann, P., (2) 384, 388 Reis, H ., (2) 214, 215, 230, 232, 236 Reisenzein, R., (1) 382, 387, 414 Reiser, B. J., (1) 192, 205 Reitman, F., (2) 363, 395 Reitzes, D. C., (1) 162, 196 Rempel, J. K., (2) 71, 142, 228, 235 Renner, B., (2) 270, 282 Rennier, G. A., (2) 118, 136 Reno, R. R., (2) 346, 388 Rest, S., (2) 222, 238, 350, 3 9 6 Rethans, A. J., (2) 93, 136 Reyes, R. M ., (2) 435, 462

Reynolds, K. D., (2) 87, 138 Reynolds, R. E., (2) 166, 180 Reynolds, T ., (2) 380, 391 Rhodes, N., (2) 81, 82, 93, 136, 141 Rhodewalt, F., (1) 127, 149. (2) 105, 129, 265, 2 86 Rholes, W. S., (1) 2, 5, 8, 29, 35, 53, 60, 66, 76, 95, 96, 127, 147, 219, 228, 256, 260, 279, 300, 310, 318. (2) 41, 62, 431, 443 , 456 Richardson, K. D., (2) 211, 232 Richman, S. A., (2) 95, 101, 135. (2) 406, 455

Riger, A. L., (2) 291, 3 2 7 Riggle, E., (1) 301, 319. (2) 165, 172, 176, 177, 179, 183, 185 Riha, J., (1) 191, 204 Riker, W. J., (2) 154, 185 Rintoul, B., (2) 317, 338 Rips, L. J., (1) 191, 195, 302, 3 1 7 Riskind, J. H., (2) 297, 3 2 7 Ritov, I., (2) 446, 447, 462 Rittenauer-Schatka, H., (1) 141, 150, 306, 320, 385, 414 Ritter, H., (1) 47, 9 6 Rittman, T. K., (2) 257, 280 Rivers, R. D., (2) 147, 149, 183, 185 Robbins, A. S., (2) 264, 285 Robbins, T., (2) 375, 388 Roberts, J. V., (2) 93, 122 Robertson, E., (2) 86, 128 Robertson, T. S., (2) 93, 122 Robey, K. L., (1) 168, 205 Robinson, G., (1) 397, 410 Robinson, M ., (2) 158, 185 Robinson, W. T ., (2) 416, 426, 462 Roche, P. A., (2) 261, 262, 285 Rodin, J., (2) 264, 271, 274, 285 Rodriguez, I. A., (1) 371, 408 Rodriguez, R., (2) 116, 123, 228, 237, 264, 28 7

Roediger, H. L., (1) 52, 96, 116, 121, 122, 148, 149, 182, 184, 205. (2) 409, 423, 436, 462 Roenker, D. L., (2) 163, 185 Rogers, A., (1) 173, 200 Rogers, C., (1) 165, 205 Rogers, E. M ., (2) 155, 185 Rogers, P. J., (1) 175, 177, 205 Rogers, R. W ., (2) 84, 102, 108, 136, 138 Rogers, T. B., (1) 161, 175-178, 188, 201, 205

AUTHOR INDEX

Rojahn, K., (2) 34, 37, 65 Rokeach, M ., (2) 80, 99, 136 Roman, R., (1) 13, 29, 30, 39 Roman, R. J., (1) 23, 3 7 Romer, D ., (2) 81, 113, 121 Ronis, D. L., (2) 97, 113, 117, 127, 136 Rook, K. S., (2) 310, 311, 335 Roper, D., (1) 370, 400, 416 Rosch, E., (1) 157, 159, 162, 203, 205, 222, 250, 281. (2) 17, 65, 416, 462 Rose, K. C., (1) 380, 409 Rose, R. J., (2) 267, 283 Rose, R. L., (2) 407, 460 Rose, T. L ., (1) 214, 278. (2) 33,34, 61 Roseman, I. J., (1) 336, 338, 343-350, 352-354, 361, 363, 365, 404, 413 Rosen, B., (2) 368, 394, 395 Rosen, J. C., (1) 187, 205 Rosen, L., (2) 377, 395 Rosen, L. D., (2) 225, 2 36 Rosen, S., (2) 84, 136 Rosenbaum, R. M ., (2) 222, 238, 350, 396 Rosenberg, M. J., (1) 42, 43, 91. (2) 71, 72, 77, 95, 136, 176, 185 Rosenberg, S., (1) 56, 96, 167, 168, 171, 197, 205, 206

Rosenblatt, A., (2) 321, 335 Rosenbloom, P. S., (1) 114, 117, 118, 148 Rosenfield, D., (1) 12, 38 Rosenthal, A., (2) 225, 236 Rosenthal, R., (1) 70, 93 Roskos-Ewaldsen, D. R., (1) 4, 5, 16, 20, 38. (2) 13, 65, 104, 107, 136, 137 Rosoff, R., (1) 388, 404 Ross, B. H ., (1) 163, 203. (2) 416, 452 Ross, J., (2) 342, 396 Ross, L., (1) 60, 96, 172, 197, 218, 233, 276, 278, 280, 284, 303, 319. (2) 87, 95, 120, 130, 131, 164, 184, 302, 306, 335, 402, 406, 408, 410, 427, 428, 440, 456, 459, 461 Ross, M ., (1) 84, 96, 191, 206, 313, 317, 319. (2) 113, 115, 131, 137 Ross, W. T ., (2) 425, 430, 462 Roth, D., (2) 297, 335 Rothbart, M ., (1) 21, 24, 38, 51, 58, 65, 74, 76, 96, 97, 135, 149, 181, 207, 212, 214, 216, 219, 223, 227, 229, 230, 234, 244, 245, 247, 250, 254, 260, 279, 281-283. (2) 5, 21, 22, 24, 27, 33-39, 49, 50, 53-55, 62, 64-66, 443, 462, 464

495

Rowe, J., (2) 270, 271, 279 Roy-Byme, P. P., (1) 370, 413 Rozin, P., (1) 188, 193, 197,206, 342, 413 Rubin, D. B., (1) 162, 205 Rubin, D. C ., (1) 189, 206 Ruble, D. N ., (1) 13, 38, 76, 96, 111, 199. (2) 35, 66, 192, 2 36 Ruble, T. L ., (1) 13, 38 Rude, S. S., (1) 372, 409 Ruderman, A., (1) 227, 283. (2) 9, 10, 67 Rudy, T. F., (2) 253, 254, 287, 291, 334 Ruehlman, L. S., (2) 299, 335 Rugg, D ., (1) 297, 319 Rumelhart, D. E., (1) 47, 96, 110, 111, 118, 148, 149, 179, 206. (2) 253, 285 Rundus, D., (1) 55, 96. (2) 416, 426, 462 Rusbult, C. E ., (2) 203-204, 227, 228, 230, 235, 2 3 6

Ruscher, J. B., (2) 46, 60 Rush, A. J., (1) 114, 145. (2) 297, 310, 328, 332

Russell, F., (2) 43, 63, 100, 131 Russell, G. J., (2) 400, 462 Russell, J. A ., (1) 323-325, 330, 359, 408, 413

Russell, M ., (1) 376, 416. (2) 291, 329 Russo, J. E., (2) 400-402, 412, 420, 430, 457, 462

Rutstein, J., (2) 217, 221, 238 Ruvolo, C. M ., (2) 3, 33, 61 Ruzzene, M ., (2) 229, 236 Ryan, C. S., (2) 25, 28-31, 62, 64 Rye, B. J., (2) 256, 282 Ryle, G., (1) 101, 149, 178, 206

s Saal, F. E., (2) 378, 392 Sabatelli, R. M ., (2) 197, 236 Sabin, T ., (1) 374, 377-379, 412 Sackett, P. R., (2) 346, 363, 366, 395 Sadler, O., (2) 87, 137 Saenz, D. S ., (2) 10, 63 Sagar, H. A., (1) 12, 38, 228, 282. (2) 33, 65

Salancik, G. R., (2) 342, 395 Saleh, W. E., (2) 291, 333 Salomon, K. L., (2) 72, 121 Salovey, P., (1) 257, 279, 373, 375-377, 413, 415. (2) 18, 21, 26-28, 63, 253, 254, 261, 264, 274, 275, 285, 287, 290, 293, 3 3 5 -3 3 7

496

AUTHOR INDEX

Saltzberg, E ., (1) 187, 205 Saltzberg, J. A., (1) 371, 416 Samuelson, W ., (2) 446, 462 Sanbonmatsu, D. M ., (1) 4, 5, 12, 18, 20, 33, 224, 282. (2) 13, 65, 70, 93, 104, 112, 117, 125, 129, 137, 402, 412, 427-430, 448, 450, 455, 458, 462 Sande, G. N., (2) 248, 249, 269, 275, 281 Sanders, G. S., (2) 268, 285 Sanders, J. B., (2) 269, 285 Sanford, R. N., (2) 145, 179 Sanft, H ., (1) 380, 409 Sansone, C., (2) 427, 429, 462 Santrock, J. W ., (1) 375, 405 Samaa, S., (2) 267, 283 Sato, K., (1) 353, 411 Sattah, S., (2) 250, 287, 407, 465 Sawyer, A. G., (2) 110, 137, 430, 463 Sawyer, J. D., (1) 80, 94 Sawyers, B. K., (2) 111, 137 Saxe, L., (2) 103, 130 Sayers, S. L ., (2) 222, 223, 231 Scandura, T. A ., (2) 340, 390 Schachter, S., (1) 11, 38, 330, 337, 351, 356, 413. (2) 220, 23 6 Schackelford, S. L ., (2) 105, 124 Scacter, D ., (1) 121, 149, 156, 178, 185, 206, 20 7

Schafer, R. B., (2) 197, 236 Schaffer, M. M ., (1) 139, 148, 162, 203, 259, 272, 280 Schaller, M ., (2) 6, 13, 63, 65 Schallert, D. L ., (1) 166, 180 Schank, R. C., (1) 44, 77, 89, 96, 133, 149, 187, 190, 192, 206, 332, 413. (2) 205, 2 3 7 Schare, M. L ., (2) 375, 413 Schavio, M. D., (1) 444, 454 Schectman, S. L ., (2) 346, 366, 390 Scheier, J., (1) 173, 200 Scheier, M. F., (1) 133, 146, 327, 332, 336, 367, 406. (2) 263-265, 277, 280, 286, 299, 329 Scher, S., (2) 96, 137 Scherer, K. R., (1) 329, 330, 337, 338, 343, 349, 351-353, 363, 365, 366, 404, 413, 414, 416

Scherf, U., (1) 380, 382, 408. (2) 217, 233 Scheuring, B., (1) 314, 315, 320 Schiffenbauer, A ., (2) 217, 2 3 7 Schilder, P., (1) 187, 20 6

Schlenk, C. T ., (1) 162, 206 Schlenker, B. R., (1) 310, 319. (2) 291, 333 Schlottman, R. S., (2) 310, 332 Schlozman, K., (2) 150, 186 Schalensee, R., (2) 417, 463 Schmaling, K. B., (2) 213, 323 Schmeer, J., (2) 363, 395 Schmidt, D. F., (2) 84, 137 Schneider, B. J., (2) 341, 395 Schneider, D. J., (1) 43, 96, 1 1 0 , 149, 248, 282 Schneider, S. G., (2) 228, 237, 264, 2 8 7 Schneider, S. K., (2) 291, 3 3 7 Schneider, W ., (1) 2, 3, 16, 17, 38, 112, 118, 119, 145, 149, 179, 206, 263, 282

Schneier, C. E ., 378, 395 Schober, M. F., (1) 290, 291, 317, 320 Schofield, J. W ., (1) 12, 38, 229, 282. (2) 33, 65 Schooler, J. W ., (1) 31, 40, 129, 145 Schoorman, F., (2) 377, 385 Schreiber, T. A ., (2) 423, 431, 461 Schroth, H. A., (2) 44, 63 Schul, Y., (2) 84, 97, 114, 132, 137, 436, 460, 463

Schuller, G., (1) 101, 147 Schultz, R., (2) 291, 336 Schuman, H ., (1) 290, 197, 316, 320 Schumann, D., (2) 402, 405,411, 412, 461, 463 Schumann, D. W ., (2) 91, 95, 100, 101, 117, 135, 137

Schvaneveldt, R. W ., (1) 163, 182, 203 Schwab, D., (2) 363, 394 Schwanenflugel, P. J., (2) 177, 2 06 Schwartz, B. L ., (1) 119, 150 Schwartz, G. E ., (2) 278, 286 Schwartz, I., (1) 9, 33. (2) 214, 233 Schwartz, J. C ., (1) 329, 335, 336, 338, 358, 359, 365, 404, 414. (2) 207, 237, 293, 329

Schwartz, M. I., (2) 219, 235 Schwartz, R. M „ (2) 291, 3 36 Schwartz, S. H ., (2) 408, 463 Schwartzman, D. F., (2) 87, 138 Schwarz, N., (1) 11, 13-15, 28, 38, 48, 141, 150, 289-291, 294, 295, 297, 299-302, 306, 307, 313-315, 317 -3 2 1 , 326, 327, 357, 379-385, 387 , 389, 393-395 , 398-401, 405, 411, 414, 415. (2) 43, 65, 70, 97, 100-102, 121, 137, 139, 165, 172, 175, 176, 183, 185, 186, 262, 286, 343, 363, 371, 373, 376, 385, 395

AUTHOR INDEX

Schulz, R., (2) 270, 286 , 291, 33 6 Schweder, R. A ., (2) 349, 359, 395 Schwinger, T ., (2) 198, 235 Scott, C. A., (2) 105, 137, 140 , 436, 437, 465

Scott, C. K., (1) 72, 9 6 Scott, W. A., (2) 80, 137 , 379, 395 Seagall, M. H ., (1) 186, 199 Seaman, S. F., (2) 317, 331 Seamon, J. G., (1) 9, 38 Searle, J. R., (1) 291, 320 Sears, D. O ., (2) 161, 186 Secord, P., (1) 172, 204, 206, 218, 229, 282 Sedikides, C ., (1) 29, 38, 53, 73, 87, 96, 97, 249, 275. (2) 25, 64 Segal, M. W ., (2) 210, 2 3 7 Segal, N ., (2) 342, 386 Segal, Z. V., (2) 297-301, 336 Seligman, C ., (2) 224, 23 7 Seligman, M. E. P., (1) 357, 391, 414. (2) 265, 285, 301, 307, 326, 334 Selman, R., (1) 172, 206 Semin, G. R., (1) 305, 320. (2) 319, 336 Semmel, A., (1) 357, 414. (2) 301, 334 Sensinig, J., (1) 168, 199 Sentis, K., (1) 154, 166, 202 Sergeant, J., (1) 186, 2 06 Servay, W ., (1) 382, 385, 414. (2) 102, 137 Seta, C., (1) 61, 9 7 Seta, J. J., (1) 12, 28, 36, 54, 61, 95, 97, 127, 131, 148. (2) 168, 184 Setterlund, M. B., (1) 168, 204 Sewell, D. R., (1) 77, 92 Sewitch, T. S., (2) 297, 336 Shacham, S., (2) 243, 245, 284 Shaeffer, D ., (2) 310, 3 3 7 Shaffer, D. R., (2) 84, 93, 116, 138, 139, 141

Shafir, E., (1) 223, 240, 241, 280 Shah, R., (2) 262, 281 Shaklee, H., (2) 443, 463 Shalker, T ., (1) 5, 35, 374, 376, 380, 397, 416. (2) 175, 182, 218, 235 Shani, A. B., (2) 357, 389 Shanteau, J., (2) 425, 464 Shapiro, D ., (1) 240, 258, 284 Shapiro, D. E., (2) 263, 286 Shapiro, J. P., (2) 291, 304, 336 Shapiro, P., (2) 44, 46, 68 Shaver, K. G., (2) 223, 2 3 7

497

Shaver, P., (1) 2, 34, 329, 335, 336, 338, 358, 359, 365, 404, 414. (2) 193, 195, 207, 208, 214, 215, 230, 234, 236, 2 3 7

Shavitt, S., (2) 13, 65, 73, 74, 86, 117, 137, 138, 179, 187, 192 Shaw, B., (1) 114, 145, 148 Shaw, B. F., (2) 297, 299-301, 328, 329, 334, 336

Shaw, L. L., (1) 326, 405 Shaw, R., (1) 186, 20 6 Shebo Van Zandt, B. J., (2) 403 , 460 Shedler, J., (2) 168, 184, 436, 463 Sheehan, M. A., (2) 158, 185 Sheffield, F. D., (2) 113, 128 Shekelle, R. B., (2) 266, 288 Shepard, R., (1) 186, 189, 206 Sheperd, J. W ., (2) 34, 63, 64 Sheppard, B. H ., (2) 449, 463 Sheppard, J. A., (2) 264, 284 Shepsle, K. A., (2) 149, 167, 181, 186 Sherer, M ., (2) 84, 138 Sherif, C. W ., (2) 79, 91, 138 Sherif, M ., (2) 8, 65, 79, 91, 138 Sherman, J. W ., (1) 74, 182, 201. (2) 27-30, 35, 44, 55, 61, 63, 65 Sherman, R. C., (2) 84, 137 Sherman, S. J., (1) 5, 12, 14, 16, 35, 38, 43, 60, 97, 100, 102, 196, 213, 219, 223, 224, 231, 233, 234, 254, 258-260, 265, 277, 278, 282, 283, 305, 306, 312, 320. (2) 3, 12, 13, 33, 61, 65, 73 , 87, 138, 167, 182, 250, 282, 341, 395, 406, 410, 420, 423, 428, 437, 443, 448, 453-456, 463 Sherman-Williams, B., (1) 221, 230, 279. (2) 33, 38, 63 Sherrod, D. R., (2) 219, 233 Shevell, S. K., (2) 191, 195, 302, 31 7 Shields, S. A., (1) 324, 414 Shiffrin, R. M ., (1) 2, 3, 16, 17, 38, 105, 114, 150, 111, 179, 205, 206, 263, 282. (2) 402, 418, 462, 463 Shimp, T. A., (2) 108, 127, 138, 139, 412, 463, 464

Shin, H. J., (1) 139, 140, 148 Shipley, E. F., (1) 240, 282 Shocker, A. D., (2) 416, 462 Shoben, E. J., (1) 42, 83, 98, 164, 177, 190, 203, 207, 208

Shoptaw, S. J., (2) 270, 28 7

498

AUTHOR INDEX

Showers, C ., (1) 134, 150 , 168, 169, 206. (2) 291, 33 6 , 343, 395 Shrauger, J. S., (2) 302, 3 3 6 Shugan, S. M ., (2) 416, 463 Shweder, R. A., (1) 174, 206, 2 0 7 Sibicky, M. E., (1) 154, 201 Sidera, J., (2) 73, 123 Sieber, W. J., (2) 261, 262, 285 Siegler, R. S., (1) 134, 150 Siemer, M ., (1) 382, 387, 414 Siladi, M ., (1) 166, 202 Silberman, E ., (1) 395, 399, 414 Silver, R. L., (2) 291, 3 3 6 Simmonds, S., (2) 101, 128 Simmons, C. J., (2) 427, 463 Simon, A., (2) 153, 157, 180, 182 Simon, B., (2) 31, 32, 65 Simon, D. M ., (2) 147, 184 Simon, H. A., (1) 107, 110, 145, 148, 327, 334, 414. (2) 403, 417, 454, 461

Simon, M ., (2) 364, 390 Simons, A., (1) 14, 15, 38, 141, 150, 306, 320, 385, 414 Simonson, I., (2) 401, 402, 412, 445, 447, 448, 463 Simpson, D. D ., (1) 181, 204 Simpson, J. A., (2) 207, 227, 2 3 7 Sims, H. P., (2) 342, 345, 362, 372, 377, 392

Sinclair, R. C., (1) 5, 36, 394, 396, 415. (2) 101, 138, 384, 395 Singer, J. A., (1) 373, 375-377, 413, 415. (2) 220, 236, 293, 336 Singer, J. L ., (1) 11, 37, 330, 337, 351, 356, 413

Singh, J. V., (2) 340, 391 Singley, M. K., (1) 113, 116, 150 Sirdeshmukh, D., (2) 99, 100, 133, 413, 460

Sivacek, J., (2) 117, 138, 408, 463 Skelton, J. A., (1) 5, 38. (2) 105, 129, 243, 250-252, 286 Skillings, J. R., (2) 249, 284 Skinner, B. F., (1) 172, 196 Skitka, L ., (1) 254, 284. (2) 95, 139 Sklar, K. B., (2) 407, 465 Skokan, L. A., (2) 271, 287, 291, 329 Skov, R. B., (1) 305, 320. (2) 33, 65, 423, 463

Skowronski, J. J., (1) 44, 51-54, 60, 76, 88, 93, 97, 215, 218, 220, 257, 268, 276, 282. (2) 256, 280, 418, 441, 453, 463, 464

Slamecka, N. J., (1) 176, 206. (2) 87, 138 Sloan, L. R., (2) 211, 212, 232, 254, 286, 428, 464 Slocum, J., (2) 380, 391 Sloman, S. A., (1) 119, 150 Slovic, P., (1) 306, 318. (2) 152, 154, 164, 182, 186, 250, 2 8 7 Slovik, L. F., (2) 203, 204, 23 6 Slugoski, B. R., (1) 303, 304, 318 Slusher, E. A., (2) 347, 396 Slusher, M. P., (1) 215, 282. (2) 37, 66, 377, 385, 395 Smari, J., (1) 166, 204 Smedslund, J., (1) 365, 415 Smith, A. F., (2) 262, 285 Smith, C. A., (1) 336, 338, 343, 344, 348-352, 354, 363, 365, 404, 407, 415

Smith, D. C., (2) 401, 461 Smith, D. E ., (2) 383, 395 Smith, E. E ., (1) 157, 159, 162, 166, 203, 204, 206, 223, 226, 240, 241, 270, 280, 403, 415. (2) 433, 464 Smith, E. R., (1) 2, 3, 6, 14, 22, 23, 26, 3 8 -4 0 , 49, 50, 53, 73, 77, 81, 90, 97, 100, 101, 115-117, 120-122, 129, 130, 135, 136, 138, 139, 143, 149, 150, 151, 219 -2 2 1 , 226-228, 230, 237, 243-248, 259, 261-263, 269, 270, 272-274, 282, 283, 285. (2) 8, 10, 11, 13, 14, 17, 19-24, 40, 66, 68, 159, 167, 172, 186, 254, 286 Smith, J., (1) 2, 36, 154, 166, 202. (2) 212, 213, 238, 302, 334 Smith, K. F., (1) 302, 319 Smith, K. H., (2) 406,464 Smith, M. B., (2) 72, 77, 124, 138, 413, 414, 464 Smith, M. C., (1) 24, 33, 51, 92, 267, 275 Smith, P. C., (2) 344,346, 395 Smith, R. A., (2) 161,183, 429, 430, 455 Smith, R. E ., (2) 408, 464 Smith, S. M ., (2) 93, 112, 135, 138, 401, 417, 452 Smith, T ., (2) 155, 186 Smith, T. W ., (1) 310, 320. (2) 264-266, 277, 280, 282, 286, 299, 336

AUTHOR INDEX

Smith, V. L ., (1) 298, 320 Smither, J. W ., (2) 346, 355, 395 Sniderman, P. M ., (2) 150, 186 Snyder, C. R. R ., (1) 2, 7, 18, 37. (2) 291, 336 Snyder, M ., (1) 134, 150 , 161, 20 4 , 229, 240, 283 , 305, 3 2 0 , 376, 415. (2) 33, 66, 73, 74, 116, 138, 159, 186, 206, 218, 228, 231, 237, 348, 349, 395, 434, 444 , 464 Snygg, D., (2) 302, 3 36 Socall, D ., (1) 69, 90, 95 Soloman, M. R., (1) 162, 206 Soloman, S., (2) 291, 336 Sommers, S., (2) 208, 2 3 7 Sonnemans, J., (1) 337, 366, 408 Sorenson, J., (2) 115, 135 Sorrentino, R. M ., (1) 132, 150. (2) 316, 331, 336

Sox, H., (2) 152, 184 Sparks, C., (1) 46, 70, 91, 93 Spear, N. E., (1) 375, 413 Spellman, B. A., (1) 242, 283 Spence, J. T ., (2) 264, 285 Spencer, S. J., (2) 96, 139 Sperber, D., (1) 71, 72, 97 Spielberger, D. D., (2) 313, 320, 3 36 Spielman, L. A., (1) 5, 14, 15, 17, 19, 26, 32, 39. (2) 300, 307, 308, 311, 314, 327, 336

Spies, K., (1) 374, 415 Spindel, M. S., (1) 336, 413 Spiro, R. J., (1) 86, 97 Spivack, G., (1) 397, 413 Spranka, M ., (1) 219, 283 Sprecher, S., (2) 197, 234, 237, 238 Spreng, R. A., (2) 99, 131 Sprinzen, M ., (1) 2, 39 Squire, L., (1) 122, 150, 185, 206 Srebnik, D., (1) 187, 205 Srinivas, K., (1) 121, 122, 149 Srinivasan, N., (2) 401, 464 Sroufe, A., (2) 192, 194, 233 Sroufe, L. A., (2) 192, 194, 238 Srull, T. K., (1) 2, 5, 8, 21, 23, 27, 39, 40, 42, 44, 47, 50, 51, 53, 56-58, 61, 64, 65, 69, 72, 76-78, 88, 90, 95, 97, 98, 100, 103, 104,112, 116, 117, 119, 129, 133, 143,144, 150, 151, 178, 181, 182, 207, 208, 212, 219, 222, 223, 228, 256,260, 2 8 3-285, 300, 301, 309, 310,312, 320, 322, 332, 334, 363, 367,374, 378, 379, 387, 415, 417. (2) 8, 34-36, 41, 66, 70, 78, 80, 83, 100, 131, 138, 141, 154,

499

161, 166-168, 170, 175, 177, 183, 186, 187, 240, 250, 288, 302, 311, 336, 343, 347, 351, 356, 373, 379, 392, 396, 397, 410, 411, 415-424, 429-431, 435, 436, 439, 440, 442, 453, 459, 464, 466

Staats, A. W ., (2) 89, 108, 138 Staats, C. K., (2) 89, 108, 138 Stacy, A. W ., (2) 291, 3 3 7 Staelin, R., (2) 400, 401, 402 Stang, D., (1) 379, 411 Stangor, C., (1) 227, 234, 269, 283, 313, 318. (2) 6, 33-37, 46, 66, 71, 139, 413, 453 Stanley, M. A., (2) 291, 338 Stanton, A. L ., (2) 258, 282 Staw, B. M ., (2) 342, 396 Stayman, D. M ., (2) 100, 101, 121, 406, 407, 433, 442, 448, 451, 453, 464 Steele, C. M ., (2) 96, 139, 291, 3 2 7 Steele, D. J., (2) 267, 284 Steele, H ., (2) 196, 234 Steele, M ., (2) 196, 234 Steenbergen, M ., (2) 176, 177, 179, 185 Steer, R. A., (2) 297, 327, 328 Stegner, S. E., I l l , 121 Stein, B. S., (1) 371, 415 Stein-Seroussi, A., (1) 18, 39. (2) 119, 139, 291, 3 3 7 Steller, B., (1) 19, 34 Stephan, C. W ., (2) 46, 66 Stephan, W. G., (2) 42, 46, 66 Stephens, L., (1) 9, 11, 3 6 Stepper, S., (1) 385, 415. (2) 108, 139 Sterling, S., (1) 243, 279 Stem, L. D ., (1) 223, 283. (2) 35, 66 Sternberg, S., (1) 176, 2 0 7 Sternthal, B., (2) 84, 114, 127, 130, 407, 418, 434-438, 453, 456, 458, 465 Stewart, D. W ., (2) 401, 455 Stewart, T. L., (1) 26, 39, 116, 117, 120, 130, 136, 143, 150, 220, 283 Stiff, J. B., (2) 107, 139 Stigler, J. W ., (1) 174, 2 0 7 Stinson, L ., (1) 76, 95 Stock, C. B., (2) 423, 463 Stokes, D. E ., (2) 165, 180 Stoltenberg, C. D., (2) 290, 3 37 Stone, T. H., (2) 347, 3 96 Stones, M. H ., (2) 291, 336 Stopeck, M. H., (2) 361, 390, 391

50 0

AUTHOR INDEX

Storms, M. D ., (1) 304, 321 Stotland, E., (2) 72, 115, 729, 130 Strack, F., (1) 13-15, 28, 38, 39, 71, 97, 100, 141, 150, 289-291, 294, 298-300, 302, 305, 306, 310, 314-316, 317, 320 -3 2 2 , 379, 385, 387, 392-395, 398, 405, 411, 414, 415. (2) 97, 100, 101, 121, 139, 187, 309, 3 3 7 Strathman, A. J., (2) 95, 101, 135, 406, 411, 455, 461

Strauman, T. J., (1) 5, 39, 169, 199, 207, 360, 361, 409, 415. (2) 258, 286, 297, 298, 331, 3 3 7 Stretton, M. S., (2) 262, 272, 285 Stroebe, W ., (2) 2, 6, 42, 60, 96, 139 Stroessner, S. J., (1) 223, 255, 283. (2) 11, 14, 15, 44, 45, 55, 61, 6 7 Stroh, P., (2) 166, 167, 169, 170, 172, 173, 111, 179, 184 Strohmetz, D. B., (1) 5, 38 Strong, K. E., (2) 82, 139 Strongman, K. T ., (1) 328, 415 Stroop, J. R., (2) 298, 3 3 7 Stuart, E. W ., (2) 108, 138, 139, 412, 464 Strube, M. J., (1) 154, 2 0 7 Suarez, S. D., (1) 172, 197 Suci, G. J., (2) 70, 134 Sujan, H., (1) 396, 397, 412. (2) 101, 133 Sujan, M ., (1) 396, 397, 412. (2) 101, 133, 412, 418, 419, 430, 431, 433, 452, 464

Sullivan, J. L ., (2) 163, 170, 1 8 4 -1 8 6 Sullivan, L. A., (2) 46, 62, 66, 71, 139 Sullivan, M. A., (1) 112, 145 Sullivan, M. J. L., (1) 370, 395, 400, 415 Suls, J., (1) 154, 207. (2) 242-246, 270, 271, 285 -2 8 7 , 314, 3 3 7 Sulsky, L. M ., (2) 345, 346, 353, 358, 359, 386, 39 6

Sun, Y., (2) 248, 281 Sundstrom, E., (2) 365, 394 Swallow, S. R., (2) 304, 3 3 7 Swan, S., (1) 69, 98 Swann, W. B., (1) 18, 39, 128, 134, 150, 305, 320. (2) 6, 33, 66, 67, 116, 119, 138, 139, 159, 186, 291, 309, 335, 337, 348, 395, 444, 464 Swasy, J. L., (2) 93, 133, 136, 139 Sweeney, J. A., (1) 193, 208 Sweeney, P. D., (1) 396, 415. (2) 310, 3 3 7

Swim, J., (2) 346, 361, 362, 396 Swinyard, W. R., (2) 408, 464 Szapocznick, J., (2) 266, 280 Szegda, M ., (1) 377, 410

T 159, Tabachnick, N., (2) 306, 3 3 7 Tabb, D., (2) 163, 184 Tafarodi, R. W ., (2) 79, 126 Tajfel, H ., (1) 9, 39, 229, 251, 283. (2) 4, 6, 8, 9, 58, 6 7 Tanford, S., (2) 102, 139 Tanke, E. D ., (1) 229, 240, 283. (2) 33, 66, 218, 228, 2 3 7 Tannenbaum, P. H ., (1) 42, 43, 97. (2) 70, 97, 134 Taraban, C. B., (2) 208, 232 Tassinary, L. G., (1) 331, 406. (2) 108, 722, 251, 252, 280 Taylor, D. M ., (2) 259, 281 Taylor, J., (2) 244, 282 Taylor, M ., (1) 227, 282 Taylor, M. A ., (2) 364, 393 Taylor, R., (1) 376, 416 Taylor, S. E., (1) 2, 17, 23, 39, 88, 94, 128, 129, 146, 151, 179, 207, 226, 221, 260, 277, 283, 304, 321. (2) 4, 9, 10, 16, 34, 50, 60, 67, 89, 112, 124, 139, 164, 186, 214, 227, 228, 237, 240, 243-246, 252, 261, 264, 271-273, 275, 277, 278, 280 -2 8 2 , 287, 288, 291-293, 302, 329, 330, 337, 338, 343, 346, 352, 363, 396, 402, 405, 407, 435, 441, 455, 464 Teasdale, J. D ., (1) 376, 380, 406, 416. (2) 307, 3 2 6 Tedeschi, J. T ., (1) 310, 321 Telaak, K., (2) 80, 125 Teng, G., (2) 86, 128 Tennen, H ., (2) 270, 271, 279 Teri, L ., (2) 297, 333 Terkildsen, N., (2) 167, 179, 185 te Schure, E ., (1) 353, 408 Tesser, A., (1) 128, 151, 260, 284, 343, 344, 350, 400, 411, 416. (2) 35, 59, 70-72, 84, 87, 88, 95, 96, 117, 133, 137, 139, 176, 184, 209-213, 225, 230, 231, 23 6 -2 3 8 , 211, 287, 318, 337, 377, 396

AUTHOR INDEX Tetlock, P. E ., (1) 29, 39, 254, 284, 350, 401, 402, 411, 417. (2) 91, 95, 139, 164, 186, 344, 373, 376, 382, 396 Tetrick, L. E., (2) 343, 391 Thagard, P. T ., 427, 433, 457 Thaler, R., (2) 152, 186, 446-448, 457, 464 Thamodaran, K., (2) 99, 100, 133, 402, 412, 460

Thein, R. D ., (1) 4, 16, 24, 25, 32, 58, 92, 117, 129, 136, 145. (2) 83, 121 Thibadeau, R., (1) 107, 151. (2) 96, 140 Thibaut, J. W ., (2) 202, 203, 228, 230, 238 Thomas, R. L ., (1) 371, 408 Thompson, C. P., (2) 163, 185 Thompson, D., (2) 363, 396 Thompson, D. M ., (1) 122, 139, 151, 204, 20 7

Thompson, E., (1) 13, 29, 30, 39 Thompson, E. P., (2) 192, 236 Thompson, K., (1) 370, 413 Thompson, L. L., (2) 214, 233 Thompson, M. M ., (2) 70, 72, 140 Thompson, S. C., (2) 223, 238, 245, 287, 291, 337, 407, 464 Thompson, T ., (2) 363, 396 Thompson, W. C., (2) 435, 441, 462 Thomsen, C. J., (2) 81, 91, 140 Thorne, A., (2) 211, 212, 232 Thornton, E. C., (2) 384, 396 Thorson, E., (2) 100, 137 Thurstone, L. L., (1) 294, 321 Tice, D. M ., (2) 87, 140 Tichener, P. G., (2) 160, 186, 187 Titchener, E. B., (2) 109, 140 Tobin, J., (2) 153, 187 Tobin, S., (2) 270, 271, 286 Todd, J. T ., (1) 186, 205 Tolle, R., (1) 392, 416 Tom, G., (2) 94, 140 Tomaka, J., (2) 72, 121 Tomarken, A. J., (1) 398, 410. (2) 292, 301, 334 Tomkins, S. S., (1) 329, 416 Tooby, J., (1) 402, 403, 407, 416 Tooke, W ., (2) 86, 128 Toppino, T ., (1) 142, 147. (2) 409, 456 Tota, M. E ., (1) 5, 6, 8, 10, 11, 25, 32, 33. (2) 8, 58, 166, 180, 300, 304, 305, 311, 312, 32 7 Toth, J. P., (1) 14, 35 Tourangeau, R., (1) 299, 321. (2) 70, 80, 140, 385, 396

501

Townsend, J. T ., (1) 101, 151 Trafimow, D., (1) 80, 81, 97 Trafton, J. G., (1) 182, 184, 201. (2) 23, 62 Traub, A. C., (1) 188, 204, 20 7 Traupmann, J., (2) 197, 234, 238 Trautman, C. M ., (2) 425, 464 Trepel, L ., (1) 243, 279 Triandis, H. C., (1) 174, 207. (2) 117, 140 Triesman, A., (1) 3, 35, 115, 147. (2) 402, 457

Trolier, T. K., (1) 213, 226, 278. (2) 2-4, 13, 42, 61 Trope, Y., (1) 24, 27, 28, 39, 40, 60, 64, 97, 128, 151, 221. (2) 98, 122, 253, 263, 269, 284, 306, 321, 421, 464 Trost, M. R., (2) 103, 140 Truxillo, D. M ., (2) 364, 388 Tsui, A., (2) 360, 396 Tuchfarber, A. J., (2) 155, 180 Tuchin, M ., (1) 298, 321 Tucker, J., (1) 353, 411 Tudor, M ., (1) 136, 144 Tulloch, W. F., (2) 92, 129 Tulving, E., (1) 6, 40, 119, 122, 139, 150, 151, 156, 178, 184, 185, 206, 207, 375, 417. (2) 434, 464 Tunis, S. L., (1) 171, 20 7 Turco, R. M ., (2) 90, 729 Turnbull, W ., (2) 86, 128 Turner, C. F., (1) 289, 316, 321 Turner, J. C., (1) 257, 283. (2) 4, 6, 9, 64, 67

Turner, T. J., (1) 77, 92, 190, 195, 329, 330, 412 Tucker, J. A., (2) 291, 338 Turk, D. C., (2) 253 , 254, 262, 285, 287, 290, 335, 33 7 Tumquist, D. C., (2) 251, 252, 280 Tversky, A., (1) 15, 36, 162, 163, 191, 208, 214, 238, 239, 241-243, 264, 266, 284, 303, 306-308, 318, 321, 334, 383, 385, 397, 410, 417. (2) 12, 24, 27, 67, 102, 129, 152-154, 182, 184, 185, 250, 283, 287, 361, 396, 409, 410, 420, 421, 424, 428, 434, 445, 447, 456, 458, 463, 465 Tversky, B., (1) 298, 321 Tweney, R. D., (2) 444, 454 Twibble, J. L., (2) 422, 459 Tybout, A. M ., (2) 92, 105, 133, 140, 406, 418, 431, 436-438, 460, 465

502

AUTHOR INDEX

Tykocinski, O., (2) 258, 283 Tyler, R. B., (1) 6, 9, 33, 37, 256, 271, 276. (2) 39, 59 Tyler, S. W ., (2) 418, 422, 465 Tyler, T. R., (2) 157, 187

u Uleman, J. S., (1) 12, 23, 24, 26, 37, 40, 51, 52, 55, 97, 98, 115, 151, 231, 263, 267, 280, 284. (2) 348, 351, 394 Underwood, B. J., (1) 189, 2 0 7 Unnava, R., (2) 91, 122, 133, 405, 411, 461, 465

Unverzagt, W. H ., (1) 56, 65, 98. (2) 372, 397

Upshaw, H. S., (1) 288, 289, 296, 312, 321, 322

Urban, G., (2) 416, 456, 465 Urbany, J. E., (2) 401, 465 Uretsky, M. B., (2) 261, 275, 281 Utne, M ., (2) 197, 234, 238

V Vaillant, G. E ., (2) 265, 285 Valins, S., (2) 244, 282 Vallacher, R. R., (1) 2, 40, 154, 208, 389, 392, 416. (2) 247, 2 8 7 Vallone, R. P., (1) 233, 284 Vanbeselaere, N., (2) 9, 67 Van Hook, E., (1) 16, 35 Van Komen, R. W ., (2) 263, 2 8 7 Vanman, E. J., (2) 46, 6 7 Vannieuwkerk, R., (2) 412, 453 Van Selst, M ., (1) 6, 3 7 Van Yperen, N. W ., (2) 272, 277, 280 Van Zandt, B. J. S., (1) 9, 36, 141, 148 Vasquez, C. V., (2) 311-315, 337, 338 Vaughn, C. E., (2) 291, 328 Venkatraman, M. P., (2) 405, 465 Verba, S., (2) 150, 186 Verette, J., (2) 203, 204, 236 Verplanken, B., (2) 113, 116, 140 Villanova, P., (2) 344-346, 386 Vincent, J. E ., (2) 219, 232, 437, 454 Vinokur, A., (2) 98, 122 Viscusi, D., (1) 398, 405 Vitousek, K. B., (2) 290, 333

Volkmann, J., (1) 295, 322 Von Baeyer, C ., (1) 357, 414 Von Hendy, H. M ., (1) 30, 34 Von Hippel, W ., (1) 212, 215, 224, 279 Von Neumann, J., (2) 445, 465 Vookles, J., (1) 360, 412. (2) 258, 283, 286 Vredenburg, K., (2) 321, 330 Vuchinich, R. E ., (2) 291, 338

w Waddell, B., (2) 199, 217, 233 Wade, S. L ., (2) 291, 334 Wagenaar, W. A., (1) 192, 208 Wagener, J ., (1) 377, 410. (2) 261, 2 8 7 Wagner, D., (1) 310, 321, 387, 414 Wahlke, J., (2) 170, 184 Walbett, H. G., (1) 353, 365, 416 Waldman, D. A., (2) 346, 366, 368, 396 Walker, C. J., (1) 260, 284 Walker, M. R., (2) 211, 212, 232 Walker, N., (2) 416, 418, 438, 465 Wall, P. D., (2) 247, 284 Wall, S., (2) 191, 193, 231 Wallston, T. S., (1) 265, 284 Walster, E., (1) 218, 276. (2) 196, 197, 220, 222, 230, 231, 238 Walster, G. W ., (2) 196, 197, 238 Walters, G., (2) 49, 62 Wan, C. K., (2) 117, 128, 245, 287, 314, 337

Wang, P. L ., (1) 185, 206 Wanke, M ., (1) 300, 321. (2) 97, 139 Ward, C. H., (2) 313, 328 Ward, D. W ., (1) 390, 391, 397, 399, 402, 411 Ward, J., (2) 416, 459 Warren, L. W ., (2) 297, 317, 338 Warren, N., (2) 302, 333 Warren, R., (2) 82, 125 Warshaw, P. R., (2) 449, 463 Wason, P. C ., (2) 444, 465 Wasserman, A. A., (2) 317, 338 Waters, E., (2) 191-193, 231, 238 Watkins, T. A., (2) 113, 129 Watson, D., (2) 265, 287, 292, 333 Watson, J. B., (1) 329, 416 Wattenberg, M. P., (2) 170, 184 Wattenmaker, W. D., (2) 21, 64 Watts, F. N., (1) 371, 416. (2) 298, 300, 302, 329, 338

AUTHOR INDEX

Watts, W. A., (2) 87, 113, 140 Watzlawick, P., (1) 72, 9 7 Weary, G., (1) 391, 392, 394, 395, 407, 409, 411, 416. (2) 290, 291, 305, 317-323, 329, 331, 332, 334, 338 Weatherford, S., (2) 148, 187 Webb, W ., (1) 76, 93 Weber, R., (1) 23, 40, 58, 93, 180, 198, 230, 284. (2) 51-53, 67, 266, 282 Weber, S. J., (2) 113, 140 Webster, J., (2) 265, 280 Wedell, D. H., (1) 297, 319 Wegener, D. T ., (2) 78, 92, 97, 102, 107, 115, 127, 135, 140 Wegner, D. M ., (1) 2, 40, 71, 98, 154, 208, 370, 389, 392, 400, 416. (2) 79, 140, 247, 248, 272, 287, 291, 338 Weigold, M. F., (2) 439, 440, 459 Wein, S. J., (2) 292, 328 Weiner, B., (1) 329, 336, 338, 344, 345, 351, 355-357, 365, 404, 411, 416. (2) 79, 140, 247, 248, 272, 287, 291, 338 Weingartner, H., (1) 370, 375, 395, 399, 413, 414, 416

Weinheimer, S., (1) 291, 302, 318, 319 Weinman, J., (2) 301, 334 Weinstein, N. D., (2) 264, 265, 288 Weiss, S., (2) 317, 338 Weissman, M. N ., (2) 292, 338 Weisz, J. R., (2) 317, 338 Weldon, D. E., (2) 84, 140 Weldon, M. S., (1) 121, 122, 149, 184, 205 Wellens, T. R., (1) 5, 3 6 Wells, G. L., (1) 254, 281. (2) 90, 93, 94, 135, 140, 402, 462 Wendelken, D. J., (2) 362, 366, 396 Wendt, S., (1) 187, 205 Wenzlaff, R. M ., (1) 71, 98, 370, 400, 416. (2) 79, 140, 310, 321, 338 Werner, H ., (1) 4, 40 Wemerfelt, B., (2) 400, 415, 416, 456 West, S. G., (2) 299, 335 Wetter, D. W ., (1) 309, 311, 3 1 7 Wetzel, C. G., (1) 11, 33 Wetzler, S. E ., (1) 193, 208, 376, 416 Wexley, K. N., (2) 363, 370, 392, 39 6 Wheeler, D., (2) 437, 454 Wheeler, L., (2) 272, 288 Wherry, M. B., (1) 168, 204 Wherry, R. J., (2) 347, 396 White, G. L., (1) 240, 258, 284. (2) 217, 221, 238 White, G. M ., (1) 330, 353, 411

503

White, L. A., (2) 346, 358, 363, 366, 367, 387, 394

White, P., (1) 376, 415 White, R. W ., (2) 72, 138, 413, 414, 464 White, S. H., (1) 156, 193, 208 Whitney, G. A., (2) 203, 204, 236 Whittlesea, B. W. A., (1) 139, 140, 151 Wicklund, R. A., (1) 10, 40. (2) 298, 329 Widaman, K. F., (2) 291, 3 3 7 Wiggins, E. C., (2) 72, 101, 122, 176, 180 Wilder, D. A., (1) 214, 229, 284. (2) 9, 31, 44, 46, 67, 68, 84, 140 Wiley, K., (2) 256, 282 Wilkes, A. L., (2) 9, 67 Wilkes-Gibbs, D., (1) 290, 31 7 Wilkie, O., (1) 223, 240, 241, 280 Wilkie, W. L., (2) 401, 425, 465 Williams, C. J., (2) 81, 125, 155, 181, 449, 455

Williams, C. L., (2) 291, 338 Williams, J. M ., (2) 298, 302, 338 Williams, J. T., (2) 147, 185 Williams, K. D ., (1) 54, 96. (2) 91, 135, 255, 281 Williams, K. J., (2) 352, 372, 374, 375, 385, 386, 387, 388, 396, 397 Williams, R. B., (2) 266, 288 Williams, R. E ., (2) 344, 388 Williams, R. J., (1) 111, 149 Williamson, D. A., 188, 208 Williamson, G. M ., (2) 199, 218, 238, 251, 281

Willis, C. E., (2) 369, 393 Willis, G. B., (2) 261, 262, 285 Wills, T. A., (2) 214, 238, 270, 288, 321, 338

Wilson, B. A., (1) 185, 201 Wilson, D., (1) 71, 72, 97 Wilson, L., (1) 190, 200 Wilson, T ., (2) 117, 128 Wilson, T. D., (1) 2, 11, 13, 31, 37, 40, 233, 245, 278, 280. (2) 70, 72, 81, 117, 140, 141, 111, 184 Wilson, W. R., (1) 2, 40. (2) 177, 187 Winkler, J. D., (1) 2, 39 Winograd, E., (1) 178, 189, 204, 208 Winograd, T ., (1) 101, 151, 178, 208 Winter, L., (1) 23, 26, 40, 51, 52, 98, 115, 151, 267, 284 Winton, W ., (1) 170, 203. (2) 18, 59 Wippnan, J., (2) 192, 238 Wittink, D. R., (2) 425, 453 Woehr, D. J., (2) 374, 377, 383-385, 392, 396

504

AUTHOR INDEX

Wolcott, D. L ., (2) 244, 248, 285 Wolff, W. T ., (1) 219, 280. (2) 429, 461 Woltz, D. J., (1) 113, 151 Wong, L. K., (2) 111, 121 Wong, R., (2) 111, 121 Wood, G., (2) 427, 4 5 7 Wood, G. C., (2) 26, 62 Wood, J. V., (1) 317, 416. (2) 214, 238, 270, 271, 273, 274, 284, 287, 288, 291, 321, 338 Wood, R. E ., (2) 347, 393 Wood, W ., (2) 81, 82, 93, 105, 106, 112, 116, 125, 129, 136, 141 Woodruff, R. B., (2) 406, 465 Worchel, S., (2) 96, 124 Word, C. O., (1) 229, 284. (2) 26, 33, 68 Worth, L ., (1) 58, 59, 94, 180, 198, 372, 389, 393, 411, 416. (2) 27-30, 36, 43-45, 56, 60, 63, 90, 100, 101, 108, 109, 131, 141, 168, 181 Wortman, C. B., (2) 294, 338 Wright, P., (2) 411, 441, 465 Wright, P. L ., (2) 86, 93, 141 Wu, C., (2) 116, 141 Wu, S., (1) 329, 359, 365, 414 Wundt, W ., (1) 323, 369, 416 Wurf, E ., (1) 154, 166, 202 Wyer, R. S., (1) 2, 5, 8, 21, 23, 27, 39, 40, 42, 44,47, 50, 52, 53, 55-59, 61, 64-69, 71, 72, 75, 77, 78, 80-82, 88-90, 9 4 -9 8 , 103, 104, 112, 116, 117, 119, 129, 133, 143, 1 4 4 , 151, 180-182, 190, 1 9 1 , 197, 198, 208, 218, 219, 222, 223, 228, 230, 256, 260, 279, 2 8 3 -2 8 5 , 300, 301, 304, 306, 309, 312, 319, 320, 322, 333, 334, 363, 367, 374, 378, 379, 381, 387, 390, 391, 397, 398, 402, 411, 417. (2) 8, 33, 34, 37, 39, 41, 44, 58, 66, 70, 75-80, 82, 97, 127, 141, 154, 159-161, 164-172, 175-177, 179, 183, 185-187, 229, 234, 240, 250, 256, 266, 288, 311, 338, 343, 347, 351, 356, 372, 373, 379, 396, 397, 410, 411, 416-418, 420, 423, 424, 426, 429-433, 435, 439, 442, 443, 453, 457, 4 6 4 -4 6 6

X Xu, G., (2) 79, 120, 409, 452

Yalch, R. F., (2) 92, 141, 402, 412, 4 66

Yalch, R. J., (1) 105, 137 Yandell, B., (2) 86, 128 Yantis, S., (2) 81, 116, 124 Yarmey, A. D ., (1) 188, 208 Yates, J. F., (2) 117, 136 Yates, S. ML, (2) 88, 92, 120, 123 Yetton, P. W ., (2) 340, 3 8 7 Yi, Y., (2) 117, 120, 418, 466 Young, B., (1) 233, 278 Young, R. D., (2) 291,332 Young, S. M ., (2) 402, 411, 461 Youtz, M. A., (2) 365, 369, 391

z Zacks, R. T ., (1) 370, 380, 409 Zadeck, S., (2) 380, 3 9 7 Zaichowsky, J., (2) 406, 465 Zajonc, R. B ., (1) 7, 9-12, 36, 37, 60, 98, 141, 151, 385, 386, 404, 412, 417. (2) 72, 109, 110, 119, 130, 133, 141, 142, 111, 187, 375, 3 9 7 Zalesny, M. D ., (2) 354, 381, 385, 39 7 Zaller, J. R., (2) 70, 142, 155, 187 Zanna, M. P., (1) 42, 94, 126, 146, 229, 284. (2) 24, 33, 39, 46, 58, 60, 68, 70-72, 75, 81, 95,100, 104, 105, 108, 112, 113, 115, 117, 118, 125, 126, 128, 131,

133, 137, 140, 142,

224, 237, 408, 436,455 Zaragoza, M ., (1) 298, 319 Zarate, M. A., (1) 22, 23, 39, 40, 49, 97, 101, 115, 138, 139, 150, 151, 226-228, 243-248, 269, 270, 272, 274, 283, 285. (2) 10, 11, 17, 19-21, 23, 66, 159, 167, 172, 186, 254, 286 Zbrodoff, N. J., (1) 114, 151 Zechauser, R., (2) 446, 462 Zehner, K. S., (2) 87, 138, 423, 463 Zeigamik, B., (1) 176, 208 Zeiss, A., (1) 374, 412 Zellner, D. A ., (1) 188, 208 Zevon, M. A ., (2) 290, 335 Zillman, D., (1) 11, 40, 385, 417. (2) 211, 235

Zimmerman, J. A., (1) 191, 204 Zimmerman, R. S., (2) 249, 253, 275, 280 Zion, C., (1) 230, 277, 364, 390 Zorich, S., (2) 384, 39 6 Zucher, H. G., (2) 157, 187 Zudney, J., (2) 33, 34, 68 Zukier, H ., (1) 402, 412. (2) 444, 461 Zuroff, D ., (2) 298, 304, 328, 338 Zuwerink, J. R., (1) 21 ,3 3 . (2) 46, 59

Subject Index

Note: This index refers the reader to pages in both volumes of this Handbook. Volume numbers (1),

(2) are given in parentheses.

A Accessibility in memory of attitudes, (1) 5, 17-21, 119-120, 125 chronic, (1) 4-5 determinants of, (1) 53-54. (2) 434-438 influence of on encoding, (1) 51-55, 127-128 on consumer product evaluations, (2) 434-442 and perceptual fluency, (1) 14-15 theories of, (1) 44-45, 47-49, 53-54. (2) 434-442 of trait concepts, (1) 4-5 , 53-54 Affect (see also Emotion) effects of on amount of information used, (1) 394-395 on attention, (1) 388-389, 392 on attributional complexity, (1) 395-396, 401-403 on cognitive capacity, (1) 388-389, 400-401 on communication and persuasion, (1) 393-394. (2) 100-102, 107-110 on elaboration, (1) 392-394 on encoding, (1) 374-375 on evaluative judgments, (1) 380-388 on information processing, (1) 388-400 research on, (1) 392-400 theories of, (1) 388-392 on memory, (1) 370-380 associative network model of, (1) 372-374

mood-congruent, (1) 374-380 resource allocation model of, (1) 370-382 state-dependent, (1) 375-376 on political judgments, (2) 174-177 on problem solving, (1) 397-399 on stereotype-based inferences, (1) 255-257. (2) 43-45 vs. emotion, (1) 326-327, 383-386 motivational influences of, (1) 377-379, 388-392 tuning function of, (1) 389-394 use of, as information, (1) 380-388, 390 about achievement activity, (1) 390 about oneself, (1) 382 as signal to danger, (1) 389 about social events, (1) 383 determinants of, (1) 386-388 Affect-as-information hypothesis, (1) 381-388 Analogical reasoning automaticity of, (1) 272-273 vs. deduction, (1) 261-263 errors in, (1) 241-243 exemplar-based processes in, (1) 243-245 processes of, (1) 235-237 role of, in social inferences, (1) 235-249 about groups, (1) 247-248 about traits and behaviors, (1) 237-249 Anxiety {see also Emotion) cognitive processes in, (2) 301-302 effects of on general beliefs, (2) 312-314 on inferential processes, (2) 314 505

506

SUBJECT INDEX

Anxiety (cont.) effects of (cont.) on memory, (2) 314-315 on perceptions of others, (2) 302-303 on predictions of future, (2) 314 schematic biases in, (2) 297-298 theories of, (2) 292-296 helplessness-hopelessness model, (2) 294-295 schema, (2) 292-294 self-discrepancy, (2) 295-296 Appraisal (see Performance appraisal) Associated systems theory, (1) 45-46 Associative network theory of memory, (1) 44-45 of self-knowledge, (1) 182-185 Attention and automaticity, (1) 3, 16-17. (2) 403-405 effect of affect on, (1) 388-389, 392 and reactions to somatic states, (2) 242-248 selective, (2) 402-408 Attitudes and opinions bases of, (2) 71-75 beliefs, emotions, and behavior, (2) 71-72 change in, (2) 73-75 functional, (2) 72-73 cognitive processes in, (2) 69-120, 161-174 consistency with behavior, (2) 116-118 definition of, (2) 70 formation of, (2) 82-95, 110-111, 161-174 strength of, (2) 80-81 structure of, (2) 75-80 expectancy-value model, (2) 77-78 probabilogical model, (2) 75-77 thought systems approach, (2) 78-79 toward risk, (2) 151-154 and voting behavior, (2) 145-146, 151-154 Attitude change (see also Persuasion) attribution processes in, (2) 104-106 cognitive response approach, (2) 85-88 consequences of, (2) 112-118 dual process models of, (1) 124-125 elaboration-likelihood model of, (2) 88-104 heuristic processes in, (2) 106-107 information-processing stages in, (2) 81-111 evaluation and yielding, (2) 85-110 information integration, (2) 110-111

reception and learning, (2) 82-85 mere exposure effects on, (2) 109-110 peripheral mechanisms in, (2) 104-110 affect-related approaches, (2) 107-110 inference-based approaches, (2) 104-106 persistence of, (2) 113-114 resistance to, (2) 114-116 Attribution dual-process model of, (1) 126 role in close relationships, (2) 220-225 role in political judgment, (2) 147-154 Attributional complexity, (1) 395-396, 401-403 Autobiographical memory organization of, (1) 192-193 Automaticity, (1) 2-31, 263-273 of analogical reasoning, (1) 272-273 and attention, (1) 3, 16-17 and awareness, (1) 7-15 of causal attributions, (1) 268-269 and controllability, (1) 28-30, 54 criteria for, (1) 2-7 decomposition of, (1) 3-7 and deduction, (1) 269-272 and efficiency in processing, (1) 3, 24-28, 114-117 and evaluation, (X) 17-21 and induction, (1) 267-269 and intentionality, (1) 16-24 models of, (1) 105-106 vs. nonautomaticity, (1) 16-24, 28-30, 115-116 properties of, (1) 3-7, 114-116 postconscious, (1) 5-6 preconscious, (1) 4-5, 7-11 of stereotype activation, (1) 21-23, 269-272. (2) 40-42 of trait inferences, (1) 23-24, 51-53, 267-268 Awareness components of, (1) 7 of influence, (1) 11-13 of mental processes, (1) 7-15 of stimuli, (1) 7-11

B Behaviors inferences of, (1) 234-235 interpretation of, (1) 51-54, 56-57, 61-62 pragmatic factors, (1) 70-72 semantic factors, (1) 51-54

SUBJECT INDEX

organization of, in memory, (1) 55-57 prediction of, (1) 231-234 Beliefs (see also Attitudes) cognitive organization of, (2) 75-78 Bias in information processing, (2) 94-98, 225-228 avoidance of in interpreting information, (1) 54 in judgment, (1) 134-136 in trait encoding, (1) 54 confirmation, (1) 305-306. (2) 249-250 corrections for, (1) 11-12, 28, 127. (2) 97 determinants of, (2) 94-98 ability, (2) 94-95 motivational, (2) 95-97 in performance appraisal, (2) 361-363 in reactions to illness, (2) 249-251 role, in attribution, (1) 303-305 “Bin” model of information processing, (1) 47-49, 104 role of procedural knowledge in, (1) 104 statement of, (1) 47-49, 181-182 Body image of others, (1) 186 of self, (1) 187-189

c Categories and group membership, (2) 8-11 primitive, (2) 7-8 role of, in inferences, (1) 253-257. (2) 431-434 Categorization (see also Social categorization) in performance appraisal, (2) 347-350 Chronic accessibility (see Accessibility in memory) Close relationships, (2) 190-230 attachment processes, (2) 190-196 in adults, (2) 193-196 in children, (2) 190-193 styles of, (2) 194-195 attribution processes in, (2) 220-225 causal, (2) 222-225 responsibility, (2) 223-224 biased information processing in, (2) 225-228 cognitive processes in, (2) 209-228 intimacy, (2) 214-216 reflection, (2) 210-214

507

social comparison, (2) 209-214 cognitive representation of, (2) 205-217 commitment to, (2) 202-209 development of intimacy in, (2) 204-216 investment in, (2) 202-209 norms, role in, (2) 196-202 communal vs. exchange, (2) 198-202 equity, (2) 196-198 need-based, (2) 198-200 role of emotion in, (2) 207-209, 216-220 satisfaction with, (2) 202-209 Cognitive complexity effects of, on performance appraisal, (2) 378-379 and vulnerability to illness, (2) 257-259 Cognitive representation of beliefs, (2) 75-78 of groups, (1) 72-76. (2) 15-25 of group variability, (2) 74-75 of health risks, (2) 259-260 of illness, (2) 252-257 influence on judgments of, (1) 59-60, 64-65, 84-89 of marital interactions, (2) 205-207 of persons, (1) 55-72 of political candidates, (2) 169-174 of self, (1) 157-165. (2) 257-259 of sensations and symptoms, (2) 251-252 of social events, (1) 76-90. (2) 205-207 of stereotypes, (1) 75-76. (2) 15-25 theories of, (1) 43-51 Comprehension (see Interpretation of information) Concepts accessibility of, (1) 4-5, 7-13 role, in organization of information, (1) 55-56 Confirmation bias, (1) 305-306. (2) 249-250 Conjunction fallacy, (1) 307-309 Consumer decision making (see Consumer in­ formation processing) Consumer information processing, (2) 399-451 attentional factors in, (2) 402-408 elaborative processes, (2) 411-415 heuristic processes, (2) 410-411 information acquisition selective attention, (2) 402-408 selective encoding, (2) 408-415 selective exposure, (2) 401-402 memory and retention, (2) 415-423 relation to judgment, (2) 420-423

508

SUBJECT INDEX

Consumer information processing (cont.) theories of, (2) 424-450 accessibility-diagnosticity, (2) 438-442 availability-falence, (2) 434-438 behavioral decision, (2) 445-448 data-driven vs. theory-driven, (2) 429-434 hypothesis testing, (2) 442-445 information integration, (2) 424-426 MODE model, (2) 448-450 omission detection, (2) 427-429 Content vs. process specificity, (1) 138-140 Contrast effects on behavior, (1) 312 on interpretation of information, (1) 11-12, 28, 54, 127 on judgment, (1) 301, 312-314, 387 Controllability of thought, (1) 28-30 Conversational norms role in interpretation of news, (1) 71-72 in person impression formation, (1) 67-70 in question answering, (1) 291-292 in response processes, (1) 290-293 types of, (1) 71 Cooperative principle in communication, (1) 292-293

D Decision making influences on, (2) 445-448 models of, (2) 445-448 Deductive inference role of, in behavioral prediction, (1) 231-235 role of group stereotypes in, (1) 225-231 Default processes, (1) 128-130 Denial role in reactions to illness, (2) 243-248 Depression {see also Emotion) cognitive processes in, (2) 298-310 effects of on attributional style, (2) 310-311 on general beliefs, (2) 304-305 on inferential processes, (2) 305-307, 319-320 on information seeking, (2) 317-319 on interpretation of social relations, (2) 309-310

on memory, (2) 311-312 on predictions of future events, (2) 307-309 motivational components of, (2) 316-320 schematic biases in, (2) 296-302 theories of, (2) 292-296 helplessness-hopelessness model, (2) 294-295 schema, (2) 292-294 self-discrepancy, (2) 295-296 Diagnosticity determinants of, (2) 441 effects on judgment, (2) 438-442 Dispositional inference, (1) 126, 131-132 Dissonance vs. self perception, (1) 126-127 Distinctiveness, role of in self concepts, (1) 170 in stereotype formation, (2) 8-14 Distributed-processing models, (1) 110-112 Dual-process models of attitude accessibility, (1) 125 of attitude change, (1) 124-125 of candidate evaluation, (2) 171-174 of consumer decision making, (2) 429-434, 448-450 of impression formation, (1) 221-222 of inference, (1) 250-253 of person perception, (1) 125-126 role of heuristics in, (1) 264-267 of self perception vs. dissonance, (1) 126-127 of social judgment, (1) 127-128, 250-253

E Efficiency of attitude change, (1) 124-125 of attributions, (1) 26-27, 126 of information processing, (1) 3, 24-28, 114-117, 128-134 of social judgment, (1) 25-26, 127. (2) 161-162 Effortfulness, role of in attitude change, (1) 124-125 in information processing, (1) 123-126 in person perception, (1) 125-126 Elaboration-likelihood model, (2) 88-104, 411-414

SUBJECT INDEX

Emotion {see also Affect; Anxiety; Depression) basic emotions, (1) 328-330 cognitive determinants of, (1) 331-368 goals, (1) 332-335 standards, (1) 335 criteria for distinguishing, (1) 330-331 definition of, (1) 324-331 vs. mood and affect, (1) 326-327, 383-386 vs. nonemotions, (1) 325-326 influence on perceptions of others, (2) 216-220 interpretation of, (2) 208-209 methods for investigating, (1) 336-337 necessary conditions for, (1) 327-328 prototypes of, (2) 207-209 role in close relationships, (2) 207-209, 216-220 theories of, (1) 337-369 action-readiness, (1) 353-355 attribution, (1) 355-358 comparison of, (1) 346-348, 361-363 criteria for, (1) 363-369 dimensional, (1) 348-351 hierarchical, (1) 338-342, 346-348, 361-363 motivation-plus-cognition, (1) 343 prototype, (1) 358-359, (2) 307-309 self-discrepancy, (1) 359-363 sequential process, (1) 351-353 Events, mental representations of marital interactions, (2) 205-207 nonprototypic, (1) 81-84 ongoing behavior, (1) 77-79 prototypic, (1) 77-81 role, in comprehension and inference, (1) 84-90 impression formation and judgment, (1) 88-89 inferences of unmentioned events, (1) 86-88 reconstruction of temporal order, (1) 84-86 theory of, 77-84 Exemplars determinants of use in inference, (1) 245-247 vs. prototypes, (1) 259-261 role in analogical reasoning, (1) 243-247

509

in conception of self, (1) 162-163 in memory, (1) 49-50

F Framing effects in consumer decision making, (2) 445-446

G Given-new contract in social communication, (1) 299-302 Groups inferences about, (1) 211-216 representation of, in memory, (1) 72-76, (2) 15-25 abstraction-based, (2) 16-20 exemplar-based, (2) 20-23 mixed models of, (2) 23-25 stereotypes of, (1) 212-216 Group variability, perception of, (1) 74-75. (2) 25-33 models of abstraction-plus-exemplar, (2) 28-30 dual-predictor, (2) 27-28 frequency distribution, (2) 30-31 PDIST, (2) 26-27 role in inference, (1) 74-75

H Halo effects on dimensional similarity, (2) 360-361 on overall impressions, (2) 359-360 on performance appraisal, (2) 358-361 Health cognition, (2) 239-279 attributional processes in, (2) 242-248 avoidance processes in, (2) 242-248 theories of, (2) 245-247 labeling, consequences of, (2) 248-251 memory processes in, (2) 261-263 representational processes in, (2) 251-257, 259-260 role of affect in, (2) 260-267, 274-278 and social cognition, (2) 240-242, 278-279 social comparison processes in, (2) 267-278 stereotype-based processes in, (2) 255-257 theoretical formulations of, (2) 274-278

510

SUBJECT INDEX

Heuristics role in attitude change, (2) 106-107 in consumer information processing, (2) 410-411, 445-446 in response processes, (1) 306-307 in stereotype-based inferences, (1) 255-257. (2) 37-40 use of, (1) 129-130 determinants of, (1) 130-134 Heuristic vs. systematic processing, (1) 124-125, 264-267 determinants of, (1) 130-132 theory of, 264-267

I Illness effects of labeling, (2) 248 mental representation of, (2) 251-255, 259-260 stereotypes of, (2) 255-257 Illusory correlations determinants of, (2) 11-14 in inferences about groups, (1) 212-213 in stereotype formation, (2) 11-14 Impression formation (see also Person memory and judgment) category-based processes in, (1) 226-227 in conversations, (1) 67-70 effects of social context on, (1) 66-70 and inductive inference, (1) 216-225 influence of event memory on, (1) 88-89 influence of traits and behaviors on, (1)

220-221 piecemeal processes in, (1) 216-217 of political candidates, (2) 165-174 theory of, (1) 61-65, 221-222 Inconsistency resolution in impression formation, (1) 57-59, 223 Induction, (1) 211-224 Inference (see also Social inference) analogical, (1) 235-249, 272-273 deductive, (1) 225-235, 269-272 heuristic processes in, (1) 129-130, 255-257. (2) 37-40, 410-411 inductive, (1) 211-225, 267-269 Information integration and attitude change, (2) 110-112 and consumer judgment, (2) 424-426, 444

theory of, (2) 111, 424-426 Information processing, models of “bin” conceptualization, (1) 47-49, 104 distributed processing, (1) 110-112, 118-119 dual-process, (1) 124-136 flow-chart, (1) 101-103, 116-117 proceduralization, (1) 105-112, 117-122 Information seeking and depression, (2) 317-319 in performance appraisal, (2) 350-351 Informativeness principle in question answering, (1) 299-302 in responses to news, (1) 71-72 In-group/out-group differences and stereotype formation, (1) 8-14 Interpretation of information pragmatic influences on, (1) 70-72 priming effects on, (1) 8-12, 51-55 role, in consumer information processing, (2) 408-414 role of procedural knowledge in, (1) 112-114

L Labeling consequences of, (2) 248-251 Loss aversion (see Risk aversion)

M Media effects and issue salience, (2) 155-157 on perceptions of importance, (2) 157-158 Memory (see also Person memory and judgment) effects on of anxiety, (2) 314-315 of depression, (2) 311-312 for information about groups, (1) 72-76 intrusions in, (1) 86-88 and judgment, (1) 59-60, 264-265. (2) 420-423 in inferences about groups, (1) 211-216 mood-congruent, (1) 374-380 for person information, (1) 51-72 procedural, (1) 49-50 role of affect in, (1) 370-380. (2) 261-263

SUBJECT INDEX Memory (cont.) social, theories of associated systems, (1) 45-46 associative network, (1) 44-45 “bin” conceptualization, (1) 47-49 exemplar-based model, (1) 49-50 schema, (1) 47 for social events, (1) 76-90 state-dependent, (1) 375-376 Misattribution of affect and emotion, (1) 14, 380-388 of arousal, (2) 220-222 Mood (see Affect) Mood-congruent recall, (1) 374-380 Mood-repair hypothesis, (1) 377-379 Motivation role in information processing, (1) 29-30, 123-136

o On-line vs. memory-based processing, (1) 211-216 Organization of information in memory (see Cognitive representation) Organizational vs. elaborative processing in self-referent recall, (1) 177 Organizational psychology and social cognition, (2) 341-344

P Perceptual fluency, (1) 14-15, 141-143 Perceptual salience determinants of, (2) 406-407 effects of on attention, (2) 406-408 on product evaluations, (2) 436, 440-441 and vividness, (2) 440-441 Performance appraisal accuracy in, (2) 351-358 definition of, (2) 352-355 and rating errors, (2) 355-357 biases in, (2) 361-368 construct validity of, (2) 357-358 contextual influences on, (2) 369-378 evaluative, (2) 370-371 interpersonal, (2) 375-378 task objectives, (2) 371-375 effects of stereotypes on, (2) 362-368

511

gender, (2) 362-365 race, (2) 365-368 halo effects on, (2) 358-361 individual differences in, (2) 378-382 and rating processes, (2) 345-347 theories of, (2) 347-351 categorization, (2) 347-350 information-seeking, (2) 350-351 training for, (2) 382-384 Person memory and judgment in conversations, (1) 65-70 effect of social context on, (1) 65-72 pragmatic influences on, (1) 70-72 processes underlying bolstering, (1) 58 evaluative encoding, (1) 56-57 inconsistency resolution, (1) 57-59 trait encoding, (1) 51-56 theory of, (1) 61-65 Person perception (see also Impression forma­ tion; Person memory and judgment) dual-process model of, (1) 125-126 Persuasion (see also Attitude change) central and peripheral routes to, (2) 88-104 cognitive responses underlying, (2) 85-88 elaboration-likelihood model of, (2) 88-104 by minority and majority sources, (2) 102-104 motivational and ability factors in, (2) 90-98 peripheral mechanisms in, (2) 104-110, 411-415 role of affect in, (1) 393-394. (2) 100-102 Politeness effect on information processing, (1) 69 role, in response processes, (1) 309-312 Political information processing (see also Vot­ ing behavior) in candidate evaluation, (2) 157-158, 165-174 categorical vs. piecemeal, (2) 171-172 effect of television news, (2) 155-158 models of, (2) 158-177 on-line vs. memory-based, (2) 173 and personality, (2) 144-145 role of affect in, (2) 174-177 salience effects in, (2) 154-157 stages of encoding and interpretation, (2) 161-162, 166-168 integration, (2) 171-174 organization, (2) 162-164, 168-171 retrieval, (2) 164-165, 169-171

512

SUBJECT INDEX

Pragmatic influences on information processing in comprehension, (1) 70-72 in person impression formation, (2) 67-72 in reactions to news, (1) 71-72 role of conversational norms in, (1) 67-72 Preconscious processes, (1) 4-5 , 115, 141-143 Priming effects on concept accessibility, (1) 7-13, 53-54 on public opinion, (2) 154-158 resistance to influence of, (1) 54 subliminal, (1) 7-13 Procedural knowledge, (1) 49-50, 99-144 change in, (1) 117-122 vs. declarative knowledge, (1) 99-112 development of, (1) 112-123 models of, (1) 101-112, 136-138 distributed-processing, (1) 110-112, 118 flow chart, (1) 101-103, 116-117 memory-retrieval, (1) 105-106, 118 proceduralization, (1) 105-112 production-compilization, (1) 106-110 von Neumann, (1) 103-105 of self, (1) 178-179 selection of, (1) 123-136 strengthening of, (1) 119-120 Processing strategies (see also Procedural knowledge) influence of affect on, (1) 388-400 Production systems compilation of, (1) 106-110 role in memory, (1) 49-50 theoretical analysis of, (1) 106-110 Prototypes (see also Stereotypes) role, in social inference, (1) 225-231, 259-261

R Relationship functioning (see Close relationships) Repetition, effects of, (2) 409 Representativeness heuristic, (1) 306-307 Responses to inconsistency in person impression formation, (1) 55, 57-59, 62 Response processes in social judgment, (1) 288-316

biases in, (1) 303-309 cognitive consequences of, (1) 312-315 contextual effects on, (1) 293-297 effects on of conversational norms, (1) 292-293 of question wording, (1) 297-302 of response scale, (1) 294-297 internal representations vs. overt responses, (1) 286-293 basis for distinguishing, (1) 289-293 role of politeness in, (1) 309-312 Retrieval processes (see also Memory) theories of, (1) 44-50, 63-64 Risk aversion in consumer decision making, (2) 445-448

s Salience (see Perceptual salience) Schemas effects of on consumer judgments, (2) 433-434 on perceptions of others, (2) 302-316 role of in anxiety and depression, (2) 292-294 in memory, (1) 47 of self, (1) 165-166 theories of, (1) 47. (2) 292-294 Selective encoding, (1) 374-375. (2) 166-168, 408-415 Selective exposure, (2) 401-402 Self (see also Self-concept) associative network model of, (1) 182-185 conceptions of context-specific, (1) 161-162 ecological, (1) 155-156 exemplar-based, (1) 162-163 probabilistic, (1) 159-160 as a proper set, (1) 158-159 as a prototype, (1) 160-161 theory-based, (1) 163-165 in a cultural context, (1) 173-175 as a knowledge structure, (1) 153-194 as a linear ordering, (1) 189-192 as a memory structure, (1) 175-179 mental representation of, (1) 179-186 perception-based representation of, (1) 186-189

SUBJECT INDEX

procedural vs. declarative knowledge of, (1) 178-179 spontaneous, (1) 168-170 Self-concept assessment of, (1) 165-171 complexity of, (1) 167-168 development of, (1) 171-173 spontaneous, (1) 168-169 varieties of, (1) 166-169 Self-discrepancy theory, (1) 359-363 and depression, (2) 295-296 and vulnerability to illness, (2) 257-259 Self-generation effect, (1) 176 Self-reference effect, (1) 175-178 Self-schemas, (1) 165-166 Social categorization automaticity of, (1) 269-271 bases of, (1) 227-229, (2) 6-11 influence on of context, (2) 9 of group membership, (2) 8-9 of stimulus salience, (2) 9-10 models of, (1) 226-227 motivation for, (2) 5-6 role, in performance appraisal, (2) 347-350 and stereotyping, (2) 4-11 Social comparison in close relationships, (2) 209-214 in coping with illness, (2) 267-274 Social inference, (1) 209-275 about groups, (1) 211-216 vs. individuals, (1) 222-225, 249-263 about political candidates, (2) 165-173 definition of, (1) 209-210 memory-based, (1) 212-213 stereotype-based, (1) 73-74, 225-231, 253-257. (2) 37-40, 362-368 types of analogy, (1) 235-249 deduction, (1) 225-235 induction, (1) 211-225 use of affect in, (1) 380-388, 390 Social judgment {see also Social inference) biases in, (1) 134-136. (2) 361-362 context effects in, (1) 301, 312-314, 387 on-line vs. memory based, (2) 420-423 of political candidates, (2) 157-158, 165-174 role of affect in, (1) 380-388. (2) 174-177, 216-220

513

Spontaneous trait encodings, (1) 23-24, 51-53, 267-268 Stereotypes {see also Social categorization) and affect, (2) 42-47 automatic activation of, (1) 21-23, 269-272. (2) 40-42 bases of, (2) 6-14 change in, (2) 47-56 evidence for, (2) 50-56 models of, (2) 49-50 definition of, (2) 1-3 formation of, (1) 211-216, 222-225. (2) 3-14 of illness, (2) 255-257 influence of on deductive inferences, (1) 225-231. (2) 37-40 on encoding and retrieval, (2) 33-34 on interpretation of behavior, (1) 21-23. (2) 33-34 on judgment, (1) 73-74, 253-257. (2) 37-40 on memory, (1) 72-76. (2) 34-37 on selective processing, (2) 34-37 use as heuristic, (1) 255-257. (2) 37-40 influence of mood on, (1) 255-256. (2) 43-45 priorities governing, (2) 39-40 Standard of comparison {see Contrast effects) Stress and coping {see Health cognition) Subliminal perception, (1) 8-13 Systematic processing determinants of, (1) 130-132 vs. heuristic processing, (1) 124-125. (2) 106-107, 411-415 nature of, (1) 132-134

T Top-down vs. bottom-up processing determinants of, (1) 253-257 Trait concepts influence of on encoding of behaviors, (1) 51-55, 61 on impressions, (1) 216-225 on organization of behaviors, (1) 55-56

u Unconscious processes {see also Automaticity)

514

SUBJECT INDEX

Unconscious processes (cont.) in comprehension, (1) 7—13 in inference, (1) 16-24

y Validity, inferences of, (2) 409-411

Vividness (see Perceptual salience) Voting behavior (see also Political information processing) attribution models of, (2) 147-151 and attitudes toward risk, (2) 151-154 economic factors in, (2) 148-154 effects of uncertainty on, (2) 151-154