Trust and Communication: Findings and Implications of Trust Research 3030729443, 9783030729448

Citation preview

Bernd Blöbaum   Editor

Trust and Communication Findings and Implications of Trust Research

Trust and Communication

Bernd Blöbaum Editor

Trust and Communication Findings and Implications of Trust Research

Editor Bernd Blöbaum Department of Communication University of Münster Münster, Germany

ISBN 978-3-030-72944-8 ISBN 978-3-030-72945-5 https://doi.org/10.1007/978-3-030-72945-5

(eBook)

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

Trust research has gained in relevance in numerous scientific disciplines over the past years. In many areas of society, it has been recognized that trust is an important basis for cooperation, social interaction, and positive relationships. It is apparently a characteristic of trust that one becomes aware of its existence and significance above all when it is endangered, eroded, or turns into mistrust. Many crises and negative developments such as polarization in some democratic societies, the emergence of populist movements, and doubts about progress in the struggle for social justice and about effective activities to slow down climate change can be understood as problems of trust. In politics, economy, culture, sports, science, and the media, organizations, institutions, and actors are concerned with how to prevent the erosion of trust and the transition from trust to mistrust, and how to develop and promote relationships of trust. If one enters the search term “trust” in databases that list scientific articles, it becomes evident that the number of publications related to “trust” has increased rapidly in recent years. If one takes media coverage as a seismograph for public discussion of relevant and current topics, the growing importance of the topic of trust becomes very clear. A glance into the archives of media confirms the picture: Journalistic articles in which “trust” appears are now much more frequent than before. Digitization, especially the Internet, plays an important role in the potential destruction of trust, as well as in its formation, stabilization, and development. Living and working under digital conditions has effects on relationships of trust. Digitization has considerably changed the context of trust relationships in many areas. Social media link skeptics and critics and perhaps lead to a loss of trust. But digital communication channels also have the potential to promote trust. When we started a research program on “Trust and Communication in a Digitized World” at the University of Münster in 2012, funded by the German Research Foundation, the digital context was still something special. Today, digitization is so ubiquitous that “the world” is digital per se, so to speak. The Research Training Group, from which the contributions in this book originate, has worked on theories, v

vi

Preface

models, concepts, and numerous empirical projects in recent years in order to explore the construction and development of trust. In four social fields—media, science, business, and sports—projects have been realized that have generated insights into trust (and distrust) at the macro, meso, and micro levels. In this book, we present some of the results and try to reflect on the implications of the findings for the domains treated. This work complements the volume “Trust and Communication in a Digitized World. Models and Concepts of Trust Research,” which was published in 2016. Bernd Blöbaum’s contribution “Thoughts on the Nature of Trust” presents and discusses models and concepts in trust research and poses the question of the theoretical potential of trust. The author deals with the concept of trust and illustrates, on the basis of representative surveys in Germany, how presupposing and volatile the measurement of trust in surveys is. Starting out from the role that models play in social science research, components are presented that are relevant to empirical research on trust. Friederike Hendriks, Bettina Distel, Katherine M. Engelke, Daniel Westmattelmann, and Florian Wintterlin discuss the methodological and practical challenges of interdisciplinary trust research. They give an overview of methodological approaches to studies dealing with trust and the prerequisites for building trust. The authors discuss the specific advantages of quantitative and qualitative research designs as well as mixed methods designs and agentbased modeling. Finally, they give recommendations for future research in different contexts. With “risk” and “transparency,” two elements of trust research that have been relatively little addressed so far are at the center of the contribution by Bernadette Uth, Laura Badura, and Bernd Blöbaum. Starting with a model that deals with the relationship between trust and risk from a process perspective and interprets transparency as a trust-building activity, the results of an empirical study are presented that deals with the perception and evaluation of different transparency strategies by media recipients. It is shown that the use of fact-checking services and the publication of editorial guidelines provide transparency of editorial work to many media users. Isabelle Freiling and Annie Waldherr ask, “Why trusting whom?” in their contribution, which deals with information processing and information evaluation in an online context. They place the evaluation of information in trust models. The authors suggest to undertake more long-term studies on trust formation and to combine the findings with network models of information diffusion. The contribution of Robin Janzik and Thorsten Quandt is also located in the “Media” section. The authors deal with the so far little explored field “trust in technology.” Based on a critical analysis of previous approaches to researching trust in technology, they develop a model of technology trust that is applied to the area of private media use from a communication science perspective. In doing so, they differentiate between social, individual, and technological influencing factors and thus complement existing research approaches. Natascha Löffler, Ulrike Röttger, and Christian Wiencierz examine the role of data-based strategic communication as a mediator of trust. In the case of nonprofit organizations, they investigate how recipients

Preface

vii

process automatic posts. From the perspective of recipients, the trustworthiness of posts is not influenced by the fact that they are created automatically. In the research field “Economy,” Viktoria Baumeister, C. Richard Hossiep, Maxine Louise Wiemann, and Gerhard Schewe investigate the role of initial trust in the context of online direct approaches via professional networks. They analyze which factors influence the trust of candidates in an online recruiter. One result of many is that candidates who generally trust other people more also trust unknown online recruiters more. Digital services are becoming more and more common in public administrations. Bettina Distel, Holger Koelmann, Florian Schmokle, and Jörg Becker examine in their contribution the role of trust for the use of such e-services by citizens. The authors provide an overview of the current state of research based on a structured literature review and present the results of an interview study from Germany. Miriam Höddinghaus and Guido Hertel discuss theoretical and practical implications for building trust under digital conditions in the area of leadership. They describe changes in the leadership context, in particular through new communication and information technology, which lead to new uncertainties and risks. Building social bonds and trust prove to be effective tools for leadership in virtual work contexts. The authors show that individuals are more likely to trust artificial intelligence for complex tasks involving computers and technology than for tasks that also require social and emotional intelligence. The paper recommends that managers must be able to manage both human employees and technology—whether as tools, team members, or colleagues. In many representative surveys, science enjoys a high level of trust compared to other social fields. Nina Vaupotic, Dorothee Kienhues, and Regina Jucks investigate how scientific literacy and science education are linked to trust. They describe which characteristics laypersons use to decide which source can be trusted. They argue that epistemic trust should be an inherent part of science education. Eva Strehlke, Rainer Bromme, Silvia Scholz, and Joscha Kärtner focus on trust decisions in the presence of information overload. Using parenting apps as an example, they illustrate how parents can make informed decisions about which programs to trust and why. Digitization is creating numerous new trust problems in the field of sports. The next three articles deal with models and methodological approaches and provide empirical findings on trust and risk in sports. With team trust, Charlotte Raue, Dennis Dreiskämper, Hannah Pauly, and Bernd Strauß analyze a hitherto little explored field. Based on the specifics of cooperation in sports teams, they adapt Shared Mental Models for this field. They show which measurement methods could be suitable for dynamic and fast game situations in team sports. The methodological implications point out the great importance of measurement methods that take into account specific situations in team sports. In their contribution, Lena Busch, Linda Schücker, Till Utesch, and Bernd Strauß deal with aspects of risk and trust when using fitness apps. They give an overview of the practice of self-tracking and address especially the usage of fitness apps. They work out the advantages and risks users associate with these apps and describe a model that includes the influence of trust in this context. The model is helpful to explain the process of initiation of, maintenance of, and dropout from fitness app usage. In competitive sports as in popular sports, the

viii

Preface

relationship between coach and athlete is always a relationship of trust. Sydney Querfurth, Linda Schücker, and Bernd Strauß explain how the coach–athlete relationship is changing through digitization. They design a model for exploring trust in coaches under digital conditions. They distinguish between trust in and trust through technology and propose specific hypotheses about the interaction and relationship of the different model elements. In the empirical part of their paper, they document that negative trust-transfer effects can occur from an untrustworthy medium to a person implementing the use of this technology. The contributions in this book provide a small extract from the research in the Research Training Group on “Trust and Communication in a Digitized World.” (Further publications can be found here: https://www.uni-muenster.de/GKVertrauen-Kommunikation/en/publikationen/index.html.) The projects funded by the German Research Foundation benefit from the collaboration of professors, postdocs, and doctoral students from communication and media studies, psychology, economics, sport science, and information science. Besides the authors, many others have contributed to this publication. I would especially like to thank Johannes Meiborg, who ensured the formal coherence of the contributions, and Andre Dechert, who, like his previous colleague Christian Wiencierz, managed the entire research program in an excellent manner. Münster, Germany

Bernd Blöbaum

Contents

Part I

Trust and Communication Under Digital Conditions as a Field of Research

Some Thoughts on the Nature of Trust: Concept, Models and Theory . . Bernd Blöbaum Methodological and Practical Challenges of Interdisciplinary Trust Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Friederike Hendriks, Bettina Distel, Katherine M. Engelke, Daniel Westmattelmann, and Florian Wintterlin Part II

3

29

Trust Research in the Field of Media

Perceptions of Trustworthiness and Risk: How Transparency Can Influence Trust in Journalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bernadette Uth, Laura Badura, and Bernd Blöbaum

61

Why Trusting Whom? Motivated Reasoning and Trust in the Process of Information Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isabelle Freiling and Annie Waldherr

83

Trust in Media Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robin Janzik and Thorsten Quandt

99

Data-Based Strategic Communication as a Mediator of Trust: Recipients’ Perception of an NPO’s Automated Posts . . . . . . . . . . . . . . . 115 Natascha Löffler, Ulrike Röttger, and Christian Wiencierz

ix

x

Part III

Contents

Trust Research in the Field of Economy

Trust Me: I Am a Recruiter—An Investigation of Antecedents and Consequences of Initial Trust in Online Recruitment . . . . . . . . . . . . 137 Viktoria Baumeister, C. Richard Hossiep, Maxine Louise Wiemann, and Gerhard Schewe The Role of Trust for Citizens’ Adoption of Public E-Services . . . . . . . . 163 Bettina Distel, Holger Koelmann, Florian Schmolke, and Jörg Becker Trust and Leadership: Implications of Digitization . . . . . . . . . . . . . . . . . 185 Miriam Höddinghaus and Guido Hertel Part IV

Trust Research in the Field of Science

Trust in Science and Scientists: Implications for (Higher) Education . . . 207 Nina Vaupotič, Dorothe Kienhues, and Regina Jucks When Play Store Knows How to Deal with Your Kid: Trust in Digital Counselling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Eva Strehlke, Rainer Bromme, Silvia Scholz, and Joscha Kärtner Part V

Trust Research in the Field of Sports

Team Trust in Sport Teams: Methodological Implications to Advance this Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 Charlotte Raue, Dennis Dreiskämper, Hannah Pauly, and Bernd Strauss Risk and Trust in Self-Tracking via Fitness Apps . . . . . . . . . . . . . . . . . . 253 Lena Busch, Linda Schücker, Till Utesch, and Bernd Strauss Trust Within the Coach–Athlete Relationship Through Digital Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 Sydney Querfurth-Böhnlein, Linda Schücker, and Bernd Strauss

Part I

Trust and Communication Under Digital Conditions as a Field of Research

Some Thoughts on the Nature of Trust: Concept, Models and Theory Bernd Blöbaum

Abstract It seems to be part of the nature of trust that it only becomes more apparent when it decreases or appears to be threatened. Based on reflections on modelling grounded in social science, this article first discusses, with reference to empirical studies (specifically: representative surveys on trust in Germany), how preconditional and volatile the measurement of trust by means of surveys is. What is measured when trust is asked for? Differences in questions, response requirements and reference points result in widely varying trust ratings. And how are the findings to be understood? Second, the paper deals with models, which support research on trust. Components of a model presented here are the trustor and trustee (with elements such as personality traits, trust and risk propensity, perception and experience) as well as the trust process—all of which are integrated into specific situations and contexts that make trust necessary. Using models of trust in online journalism and in sources as examples, the references of trust models are described. Finally, the theoretical capacity of trust is viewed sceptically. Implications for empirical research on trust—which needs to be rethought in light of trust concepts, models and theories—are discussed in the conclusion. Keywords Trust · Empirical trust research · Trust models · Media trust · Concept of trust · Theory of trust

1 Introduction Trust is not visible and thus difficult to describe. Perhaps it is detectable in a close relationship—but often only when it is in danger of fading away. Trust exists, so it is something tangible. But at the same time trust remains abstract. Something that is

B. Blöbaum (*) Department of Communication, University of Münster, Münster, Germany e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 B. Blöbaum (ed.), Trust and Communication, https://doi.org/10.1007/978-3-030-72945-5_1

3

4

B. Blöbaum

present or should be present in society as a social community, something that characterizes the quality of personal relationships. It seems that trust in the private sphere as well as in the public perception is above all always addressed when it is violated, destroyed, reduced and burdened. It seems that crisis is a close associate of trust. The existence and importance of trust come to the attention of the public especially when familiar structures, established procedures and rules, traditional expectations and proven behaviour patterns are radically changed or threatened. Not every crisis is a crisis of trust. But it is striking that a dominant context of trust is the crisis motive. The value and significance of trust often reveal themselves in the context of a crisis. Many crises can be interpreted as crises of trust. From the Cuba Crisis in 1962, which involved a conflict between the USA and the USSR over the stationing of missiles, to the oil crisis in 1973, which brought Germany car-free Sundays, to the financial crisis in 2008, in which states had to save banks from collapsing, and to the current crisis of the European Union after the Brexit and the Corona crisis: trust has been called into question everywhere—between states, regions, economic actors and individuals, in politics, business and science. Do crises arise from a lack of trust or does a lack of trust arise from crises? Therapists (in partnerships and families) and diplomats (in political conflicts) advise confidence-building measures when symptoms of a crisis arise. Building and restoring trust are seen as effective means of overcoming conflicts. In diplomacy and therapy, priority is given to communication for crisis management and trustbuilding. Trust and communication are closely related to each other: through communication, trust can be established and maintained. Trust is a social phenomenon linked to communication. Trust, as Hartmann (2020) points out, has become such a natural part of our environment that we only realize it was there when it erodes, dries up or turns into distrust. It seems to be a paradox: only the disappearance brings trust to the fore. We apparently become more conscious of trust when trust itself becomes less present. Hartmann (2020) makes an analogy between trust and the air: the true value of air is only revealed when it becomes too thin, too dirty, too scarce. Following this analogy, it should then also be possible to measure trust in the way air density and air quality can be measured. Thus, trust becomes an empirical quantity—if not directly visible, then at least describable. This article, as well as the other contributions in this book, deals with trust as an element in society and in some of its sub-segments such as media, sports, science and business. The structure and argumentation of the text are based on studies grounded in the discussion on model building and theory development in social science. It begins with the question of what the object, phenomenon, context or field to be described and represented by means of a model or theory is. How is trust conceptually conceivable? A reflection on the concept of trust is followed by a presentation of models before finally asking about the theoretical capability of trust. The considerations will build upon reflections on key factors in the trust process (Blöbaum 2016) and are based on discussions and findings from the Research Training Group “Trust and Communication in a Digitized World” funded by the German Research Foundation (DFG) from 2012 to 2021.

Some Thoughts on the Nature of Trust: Concept, Models and Theory

5

2 Trust: Concept and Function Hardly a week goes by without the publication of a recent trust survey, the findings of which are reported in the media. The editorial programmes of mass media are highly sensitive to crises. Data on the distribution of viewpoints on trust in politics, politicians or political institutions therefore easily find their way into reporting. In contrast to surveys that determine intended voting behaviour, surveys on trust provide a kind of diagnosis of the emotional state, proximity and distance to the objects surveyed, be they political or economic actors, political organizations such as the EU or institutions such as the German Federal Constitutional Court. If the media are concerned about trust, this not only documents an editorial preference for this topic. In their reporting, media also reflect which topics are relevant to society and from which perspectives. A very superficial search shows the growing importance of the issue of trust—or at least that topics are increasingly being presented in the media in the context of trust. Before 1970, the search term “trust” in the German news magazine Spiegel (print and online) returned 2412 results, before 2000 the statistics showed 10,198 hits and since 2005 the number of hits for “trust” has exploded to 28,778 (https://www.spiegel.de/, retrieved on 29 July 2020). Anyone who uses surveys to ask about trust in different objects learns something about society. But what? A study by the Bertelsmann Stiftung bears the alarming headline “Dwindling Trust in Politics and Parties. A Threat to Social Cohesion?”1 (Faus et al. 2019) Based on a representative survey conducted in Germany in 2017, the analysis states that 25.2% of respondents have no or very little trust in the federal government, 29.6% have a great or very great deal of trust in the federal government and almost a half of the respondents (45.2%) answer that it depends (Faus et al. 2019, p. 72). In the same survey, 57.3% agree with the statement “fairly” or “completely” to the item: “All in all, I am satisfied with the democracy as it exists in Germany”. Only 13.7% agree “not at all” or “little” with this statement (Faus et al. 2019, p. 44) The questions about satisfaction and trust identify attitudes and feelings. But what exactly do they measure? What does someone have in mind when answering the question regarding satisfaction with democracy? Does the answer reflect how happy someone is with democracy in Germany compared to Finland, the USA or Hungary? Does the answer to satisfaction relate to the possibility of being able to participate in elections in general? Does one value the balance of power or the federal system? Or does one evaluate the possibility of citizen participation in urban development plans? Based on the answers to the question of trust, it also remains unclear whether a respondent judges the federal government in relation to his or her state or local government, whether he or she trusts individual politicians or whether trust relates to specific administrative decisions. Furthermore, surveys on trust often suffer from the problem of unclear reference points: What is the reference point of trust for respondents? 1

Here and for the following German text excerpts: Author’s translation from the German.

6

B. Blöbaum

The difficulty is not only to determine what the data from empirical studies of trust mean. The interpretation of the findings is equally difficult. Is it a matter of concern when 25.2% of citizens in Germany have little or no trust in the federal government? It is a very plausible result that half of the respondents stated “it depends” when asked for their amount of trust. It expresses in a certain way that there will be periods and matters when people either trust or do not trust their government. The question remains whether the issue of trust in government is an indicator of political approval or disapproval rather than a valid indicator of trust. In trust research, some authors reflect on what trust surveys measure. “Whether citizens would be prepared to say that they trust their government will be largely dependent not only on perceptions of the legitimacy or rightness of political rule, but more directly on the governments´ performance in the provision of wellbeing, social welfare, or some other good that might result from practices of policies of the institution in question”, Barbalet (2019, p. 18) assumes. With regard to studies on media credibility, Meyen (2020) concludes that these surveys would measure an “artefact”. The answers would “at best express satisfaction with the social system as a whole” (Meyen 2020, p. 60). Uslaner (2018, pp. 9–10) sums up: “Trust in political institutions depends heavily on the state of the economy”. There is a broad consensus in social science that trust is essential for the functioning of modern societies. “In late modernity trust becomes important for the maintenance of social order by preserving the viability of social relationships” (Barbalet 2019, p. 13). The increase in complexity in the course of the transformation of modern society is seen as a driver for the growing importance of trust (Luhmann 1979). The historian Frevert (2013) describes how trust from dyadic relationships extended to larger social units such as business and politics. Simmel (1978, 2008), whose studies mark the beginning of sociological research on trust, understands trust as a necessary mechanism for social interactions under the condition that individuals know each other less and less and do not know much about each other. “Without the general trust people have in each other, society itself would disintegrate” (Simmel 1978, pp. 178–179). The establishment of trust as a mechanism for reducing complexity is also the central theme of other social theorists such as Luhmann (1979) and Giddens (2013). Individualization (Adam et al. 2000) on the one hand and the increase of complexity on the other hand are repeatedly highlighted in studies as reasons for the importance of trust (Barbalet 2019). “Without trust, societies really could not exist” (Sasaki 2019, p. 1). Research on trust is expanding. This is an indicator of the growing importance of trust research in science. At the same time, it can be understood as a reaction to processes of social change, which are being described with terms such as globalization and speeding-up. Barbalet (2019) describes how, from the 1950s onwards, more and more scientific disciplines turned to trust as a research topic: economists, lawyers, psychologists, management researchers, sociologists. Counting English contributions with the word “trust” listed in Google Scholar (as of February 4, 2019) Barbalet (2019, p. 12) documented the enormously increased scientific interest in trust: In the decade from 1900 to 1909 13,700 contributions to trust are listed, 1950 to 1959 there were 22,600 articles, 1970 to 1979 already 136,000 and

Some Thoughts on the Nature of Trust: Concept, Models and Theory Table 1 General Trust in the Media

“In general, one can trust the media” Fully agree Rather agree Partly agree/partly disagree Rather disagree Fully disagree

7 2019 5 22 37 22 13

Figures in percent; n ¼ 1017 (IfK Trendstudie 2018, 2019)

from the 1990s on there was a rapid increase: 1990 to 1999 there were 1,460,000 contributions and in the following decade 2,030,000 articles. This data supports the assumption that the experience of increasing social complexity promotes trust research. The growth and differentiation of trust research have contributed to a better understanding of the elements that contribute to building and losing trust. The concept of trust has become more multifaceted. The problem of describing trust as an empirical quantity still exists. This also involves basic questions such as whether trust is more cognitive, affective or emotional in nature (PytlikZillig and Kimbrough 2016). It has already been pointed out above that in surveys it is not clear what respondents refer to when answering questions related to trust. Measurement difficulties will be illustrated here using an example from our own research on media trust. In representative surveys in Germany, between 2017 and 2019 we interviewed around 1000 citizens aged between 14 and 64 years living in private households with Internet access. The study (cf. Blöbaum 2018) shows that the findings and their interpretation depend, among other things, on the question posed, the answers given (scaling) and the specified reference points for media trust. Question: In all three surveys, the interviewees were asked to agree with the statement “In general, one can trust the media”. For 2019, the answers were distributed as follows (Table 1): Compared to 2017, the numbers have hardly changed, they are stable. The data shows that slightly less than a third of those surveyed consider media to be trustworthy, slightly more than third do not trust the media and another third have a differentiated view on this matter. With all answers, it is not clear what a respondent is referring to when he or she states his or her attitude of trust towards media: Does he or she reflect his or her attitude towards media in comparison to his or her attitude towards politics, business, justice or science? Does he or she read, listen or view many different media and perhaps has a cross-section of the media in mind? Does he or she refer to his or her daily newspaper, the online offering of a news magazine, a news portal? Is his or her concern about reporting on a specific topic or field (sport, migrants, social justice, etc.)? Does someone think of journalists in general or of a specific journalist when he or she comments on media trust? For those who partly agree and partly disagree, it may be assumed that they have a more differentiated view. In some situations, on some topics, with some media, and some journalists, etc. they tend to trust more—and in others they tend to trust less. This nuanced attitude towards media in matters of trust has at least a certain plausibility.

8

B. Blöbaum

Table 2 General trust in media—variation of response specifications

“In general, one can trust the media” Fully agree Rather agree Partly agree/partly disagree Rather disagree Fully disagree

2019 5 22 37 22 13

2018 4 29 41 24

Figures in percent; 2019 n ¼ 1017; 2018 n ¼ 1020 (IfK Trendstudie 2018, 2019) Table 3 Media trust and media distrust—variation of the questions and answers “In general, one can trust the media” Fully agree Rather agree Partly agree/partly disagree Rather disagree Fully disagree

2019 5 22 37 22 13

“How would you generally describe your attitude towards the media?” High level of trust Medium level of trust Low level of trust

2019 4 24 19

Neither trust nor distrust Low level of distrust Medium level of distrust High level of distrust

19 13 10 11

Figures in percent; n ¼ 1017, 1018 (IfK Trendstudie 2018, 2019)

Response specifications and scaling: As already mentioned, the representative survey shows that there are stable values for media trust in all three years from 2017. What happens if the question of general trust in the media remains the same, but the response specifications change? Table 2 documents how the picture of media trust then changes. When respondents no longer have the opportunity to choose “partly agree/party disagree” as their answer, the proportion of those who say they rather or full disagree that, in general, you can trust the media increases significantly from 35% to 65%. Instead of one-third, the proportion of media sceptics in Germany is then two-thirds, which is much more worrying from a media perspective. If respondents are forced to take a clear stance on media trust by giving corresponding answers, those who answer in a differentiated manner here evidently tend to take a media-sceptical stance. The central importance of the questions posed and the answers offered is illustrated by another finding from the representative surveys in Germany. In 2019, media trust was surveyed in two different ways: as agreement with the sentence “In general, you can trust the media” and as an attitude towards the media with answers ranging between “high trust” and “high mistrust” (Table 3). If “trust” as a category shifts from the question to the answers, a surprisingly modified picture emerges, especially among those who are weighing up the issues. While 37% state that they partly agree and partially disagree that you can trust in the media, only 19% of the same respondents qualify their attitude as characterized

Some Thoughts on the Nature of Trust: Concept, Models and Theory

9

Table 4 Reference points of media trust Item In general, I trust the reporting on political issues in the media In general, I trust the reporting on sport issues in the media In general, I trust the reporting on climate change in the media In general, I trust my regional/local newspaper In general, I trust journalists

Fully/rather agree 52

Rather/fully disagree 46

74

20

50

47

65 44

31 53

Figures in percent; n ¼ 1014, 959, 996, 988, 992 (IfK Trendstudie 2019)

neither by trust nor distrust. The given possibility of being able to choose between trust and distrust among seven levels results in a differentiated picture: half of the respondents position themselves in the middle, namely as having low trust, low distrust or neither trust nor distrust. The proportion of those who position themselves in the middle is significantly lower for the question with seven possible answers than for the question with five possible answers. While the response behaviour with regard to the negative options (rather or fully disagree regarding media trust and different levels of distrust) is consistent with one-third each, the proportion on the positive side increases from 27% to 47% of respondents who have at least “low” but predominantly “medium” trust. Reference point of media trust: What do respondents refer to when they give details of their media trust? In order to provide some guidance on how to answer this question, various specific references of trust in the media sector were given in the 2019 survey (Table 4). Without going into all the details and implications at this point: The values documented in the table tend to show that the trust ratings are better when specific reference points are asked for. In 2019, a good quarter of respondents trusted the media in general, but more than half (53%) have trust in political reporting, and two-third trust their regional or local newspaper. Within the thematic areas covered by the media, the approval ratings also vary. There is considerably more trust in sports reporting than in political reporting, which is considered as controversial as reports on climate change. In the very general question about trust in journalists, the percentage of positive answers is again significantly lower than in the questions about more specifically named objects of trust. The variance of the data documents that answers to trust questions apparently reflect attitudes and opinions of the respondents. But is this already trust? Below, trust is understood as a relationship construct, as a process that involves an action associated with risk. The relationship, action and risk components can hardly be depicted in trust surveys. The presentation of results from representative surveys of the population on media trust makes it clear that the analytical substance of the concept of trust—at least for the respondents—is apparently not clear. Measurement problems and thus a certain volatility of empirical trust research become obvious. Trust remains an

10

B. Blöbaum

“empirically difficult phenomenon to access” (Hartmann 2001, p. 8). A diagnosis in which survey data is used to draw conclusions about trust crises should at least be critically examined. Instead of interpreting quite small differences between repeated trust surveys as a loss (predominantly) or gain in trust, many measurements could also be seen as confirmation of stable trust relationships. Hartmann (2020, p. 85) criticizes that trust in surveys is reduced to attitudes and that the specific contexts of action are not taken into account. He speaks of “questionnaire-generated trust” (p. 28) and doubts whether basic trust can be measured at all (p. 50). The question of generalized trust has been accompanying research since at least the 1940s. The standard question reads: “Generally speaking, would you say that most people can be trusted or that you need to be very careful in dealing with people?” Bauer and Freitag (2018, p. 8) state: “Despite some innovations, researchers today primarily use modifications of questions that were introduced in the 1940s and 1950s for social trust and in the 1960s for political trust”. General trust is seen, in a sense, as a glue in a socially and spatially differentiated society. But does self-disclosure about whether one can trust others or if one cannot be careful enough say something about trust as behaviour? In surveys on general trust, the concept of trust often appears emphatically loaded. Sasaki (2019) calls for better research into the structure and consequences of trust, linking the micro, meso and macro levels, comparative, quantitative and qualitative research and long-term studies. “So far as longitudinal studies are concerned in a strict sense researchers are able to identify both stable and unstable (changing) components of people’s trust” (Sasaki 2019, p. 6). However, if trust is not only understood as a mood, but as an element of a process that leads to specific decisions and actions, then it makes sense to include the underlying specific situations and contexts in the research: “Trust is situationspecific” (Bauer and Freitag 2018, p. 2). At the very least, in addition to general questions of trust, questions relating to specific situations and reference points of trust should also be asked. This breaks down the concept of trust into two components: general trust as an attitude and specific trust in the context of action and decision-making. A static concept of trust operates in an individual-centred way: “Trust operates in terms of dispositions, beliefs or cognitions and feelings or affects or emotions, and these are always properties of individual persons” (Barbalet 2019, p. 13). In fact, it seems to be beneficial to think of the trustor as an individual, simply because trust presupposes a free will. But as soon as an individual enters into a relationship of trust, as soon as someone accepts the risk of being vulnerable, trust becomes a process. It starts out in a state, but then unfolds a dynamic that is geared towards achieving a goal and thus leads to change. This change can also consist in stabilizing the notion of trustworthiness. The conceptual and analytical vagueness of trust is complemented by practical measurement problems, which are now being intensively discussed in trust research. Bauer and Freitag (2018) point out that it is unclear what respondents associate with “most people” when asked about generalized trust. Under the heading “Are surveys on trust trustworthy?” Miller and Mitamura (2003) draw attention to

Some Thoughts on the Nature of Trust: Concept, Models and Theory

11

misinterpretations in the measurement of generalized trust, particularly because of the choice of words for the question, which leaves room for interpretation, and also because of cross-cultural comparability. Delhey et al. (2011) also state that although “most people” is predominantly associated with out-groups, there are considerable national differences in the understanding of how broadly the circle of out-groups can be understood. (On methodological problems in trust research, see also the contribution of Hendriks et al. in this volume; Engelke et al. 2019; Fisher 2018; Bachmann and Zaheer 2013; Dietz and den Hartog 2006). It was mentioned above that trust is often on the agenda in crisis situations. A common narrative is to give negative connotations to dwindling trust. But why is less trust more harmful than greater trust? Even if it seems counter-intuitive at first: If one relates trust to the complexity of society, one could argue that increasing complexity creates a greater need for trust. Low levels of trust would then indicate a deficit in complexity management. When assessing whether a lot of trust is good and little trust is bad, it is also important to keep the specific situational reference of trust in mind. Of course, a high level of trust in politicians seems desirable in a democracy—but is it not also a sign of an enlightened public sense of social responsibility if trust in a politician who makes decisions at the expense of the general public, who, for example, restricts the independence of the judiciary, decreases? In democracies, for example, a loss of trust in a political party that no longer receives one’s vote is a central element in maintaining a liberal structure. Withdrawal of legitimacy through voting in the event of a loss of trust is a constitutive element of democracies. Petitions, demonstrations, protests, media with a critical and controlling capacity towards politics—there are many institutionalized ways to articulate distrust without endangering the system itself. “Democracies institutionalize distrust ideally by containing and channelling distrust in ways that keep distrust from generalizing, especially to those parts of government with mandate and capacities to provide generally agreed goods for society” (Warren 2018, p. 5). Irrespective of these considerations, it remains unclear in the case of general trust issues whether the attitudes expressed in these questions are also relevant to behaviour. Not trusting the media might not necessarily lead to an abstinence from media use—but perhaps to a change in the media repertoire that is used. The discussion of difficulties in the empirical operationalization of trust should not be an argument against surveys or other research methods in trust research. The aim is to point out the existing analytical ambiguity of the concept of trust and to critically examine the scope of research results.

3 Trust: Model(s) In science, models are a tool to improve the understanding of phenomena and processes. Models are idealized forms “whose representation is based on a wellconsidered simplification” (Saam and Gautschi 2015, p. 27). There are no generally valid rules for modelling. The “Encyclopedia of Philosophy and Philosophy of

12

B. Blöbaum

Science” (Mittelstraß 2016, p. 423) describes models as “idealizing reduction to relevant features” and “representation of unclear specific or abstract objects or facts”, with “[t]he representation of object-like components in many cases taking second place to the representation of their relational-functional relationships (structure)”. Models have a “didactic function” (Mittelstraß 2016, p. 424) and serve to provide simplification. As a rule, they are composed of three components: Elements, relations of these elements and their operations (Saam and Gautschi 2015). “The characteristic [...] that all theoretical models have in common is that they provide representations of parts of the world, or the world as we describe it” (Hughes 1997, p. 325). In the literature on modelling in (social) science disciplines, concepts, models and theories are sometimes separated, but sometimes these terms are used synonymously. For Jaccard and Jacoby (2010) concepts are preliminary stages to models. They are “building blocks for all thinking” and “generalized abstractions” (Jaccard and Jacoby 2010, p. 11). The conceptualization of the scientifically considered phenomenon precedes the modelling. In their guide to model and theory building (for the authors the terms theory and model are “interchangeable” (Jaccard and Jacoby 2010, p. 29)), Jaccard and Jacoby state that “the distinction between theories and models in the social science literature is not always apparent” (p. 28). Models are considered as a “special type of theory”, “portions of theories”, models “are derived from theories”, “simplified versions of theories”, and “models represent correspondence between two or more theories” (p. 28). For the authors, a model or theory is “a set of statements about the relationship(s) between two or more concepts or constructs” (Jaccard and Jacoby 2010, p. 28) “A theoretical expression refers to any external symbolic representation of an internal conceptual system” (p. 30). In the following elaborations, the concepts of trustor and trustee will be discussed in order to contextualize trust and trust as a process and to develop a model of trust. To relate these different elements to each other describes the model building. If trust is modelled as a process, this does not imply that all elements constituting or influencing trust should be included. There are no general rules for modelling (Morgan and Morrison 1999). As in other idealized models, only those features that are considered relevant are captured. This reduction of complexity for the purpose of illustration also affects the elements. The model does not aim to fully describe the relationships of the elements that can be used to represent trust. The model primarily serves the purpose of simplification and can, in the best case, help to support trust research in structuring and thus providing a better understanding of the trust process. It is descriptive rather than explanatory (Saam and Gautschi 2015, p. 29). Models map a terrain (Giere 1999)—and as with maps, the representations can be at different scales. Before the model is presented, an attempt is made to name the most important components of the trust process—as they appear to the author. The following presentation builds on the considerations of key factors in the trust process (Blöbaum 2016). Mayer et al. (1995) published a model of trust that has been well-established for many years. This model may be based on what Jaccard and Jacoby (2010, p. 31) call “consensual validation”, a model whose utility has been proven over the years in

Some Thoughts on the Nature of Trust: Concept, Models and Theory

13

Proposed Model of Trust Factors of Perceived Trustworthiness Perceived Risk

Ability

Benevolence

Trust

Risk Taking in Relationship

Outcomes

Integrity

Trustor s Propensity

Fig. 1 Proposed model of trust. Source: Mayer et al. (1995)

research on trust in individuals, groups and organizations (Fulmer and Gelfand 2012; Schoorman et al. 2007; Schoorman et al. 2015). Trust is defined as the “willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party” (Mayer et al. 1995, p. 712) (Fig. 1). The model does not explicitly visualize the parties (trustor and trustee involved). The trustor is reduced to his tendency to trust. On the side of the trusted party, three test areas for assessing trustworthiness come into focus: Ability is “the group of skills, competencies, and characteristics that enable a party to have influence within some specific domain” (Mayer et al. 1995, p. 719). Benevolence refers to the “extent to which a trustee is believed to want to do good to the trustor, aside from an egocentric profit motive” (Mayer et al. 1995, p. 718). Integrity is understood as the “consistency of the party’s past actions, credible communication about the trustee from other parties, belief that the trustee has a strong sense of justice” (Mayer et al. 1995, p. 719). Whether or not an act of trust as taking the risk then occurs is the result of a quasi-empirical balancing procedure, in which the perceived risk is related to the benefit gained from trust. “We propose that the level of trust is compared to the level of perceived risk in a situation. If the level of trust surpasses the threshold of perceived risk, then the trustor will engage in the RTR (risk taking in relationship, BB). If the level of perceived risk is greater than the level of trust, the trustor will not engage in the RTR” (Mayer et al. 1995, p. 726). If the risk assessment produces a positive result, the perceived trustworthiness of the trusted party improves; in the worst case it decreases. The integrative model of organizational trust presented by Mayer et al. (1995) contains many key ideas of trust research. Trust is described as a process and the

14

B. Blöbaum

situational character is taken into account. The role of risk is included as well as factors that influence the rating of trustworthiness. The elements ability, benevolence and integrity have proven to be helpful variables in the trust process in further research. It can be regarded as a consensus in trust research that trust is a relationship construct, between trustor and trustee. The trustor is usually imagined as a person. For Sasaki, “trust relations are essentially dyadic, between two individuals, even if one of those ‘individuals’ is a collective entity in the form of an institution, such as money or the government, or an organized body such as a profession” (Sasaki 2019, p. 13). There is no agreement in the scholarly literature on whether the object of trust can also be only an individual. “Only persons as individual actors can provide or reciprocate trust, or be the objects of trust”, argues Barbalet (2019, p. 14). In many concepts, trust can be directed not only at individuals but also at collective actors such as organizations and institutions, as well as at systems such as entire social sectors like politics, the economy, science, etc. In this contribution the view is preferred that groups, communities, organizations, institutions, professions and social systems can also be receivers of trust—and not only their individual representatives as access points (Giddens 2013) for trust. Trust is always associated with the risk of disappointment, which makes the trustor vulnerable. However, risk is not part of all trust concepts and it is also unclear whether the risk must be perceived. (On consensus and conflicts in trust research, see the helpful overview by PytlikZillig and Kimbrough (2016).) The example of risk can be used to illustrate how contingent the formation of trust models is. For Mayer et al. (1995), risk perception is part of their model and thus becomes an empirically testable element. If the (calculatory) estimation of the risk of disappointment played no role in the trust process, risk would not have to be integrated into a model. For Luhmann (1988), the state of confidence exists when there is no risk perception. Möllering (2006, p. 7) regards trust as a state of positive expectations of the trustee, “irrespective of whether the trustor is conscious of this or whether it is directly observable by others in any way”. If one includes risk in a trust model, the question remains whether this risk is a consciously integrated element in the trust process or an inevitable background phenomenon linked to trust, which is not included in the assessment of trustworthiness but is only realized after the trust process. In addition to the “whether”, it is therefore also a question of “where” risk is implemented in the model. Models should help to structure the research by representing elements and their relationships. Taking risk as an example, it can be shown that a model is by no means the result of empirically detected relationships between elements, but that the choice and arrangement of these elements can steer research in one direction or another. If risk is not seen as part of a trust process, risk does not have to be part of the model. If risk plays a role in the trust process, it must be built in as an element. Risk in the form of a personal risk propensity of the trustor or as a balancing of, for example, ability, benevolence and integrity prior to a trusting action becomes an empirically measurable variable. Empirical trust research must then take the risk component into account.

Some Thoughts on the Nature of Trust: Concept, Models and Theory

15

There is broad consensus that the beginning of a trust process is linked to a specific objective. The initiation of this process is entirely voluntary. Trust cannot be forced, neither on the part of the trustor nor on the part of the trustee. The act of trust is directed towards the future. However, it is influenced by past perceptions and experiences. How these experiences develop on the side of the trustor is discussed in detail in trust literature. The development of a general propensity to trust as a personality trait is often associated with early childhood experiences (Erikson 1993). Overall, however, little research has been done on socialization effects for the development of trust propensity. There is no consensus on which factors influence the trust process and to what extent. On the one hand, these factors are linked to the personality of the trustor, including his or her general disposition to trust and take risks. On the other hand, they are linked to characteristics on the side of the trusted party that have the potential to be decisive for the assessment of its trustworthiness. This includes competence and the evaluation of the party’s intentions. First-hand and secondhand experience with the trustor and its perception can be included in the assessment of trustworthiness (Blöbaum 2016; Sztompka 1999). The many facets of trust can hardly be integrated into one model: Trust as belief, attitude, intention, behaviour. PytlikZillig and Kimbrough (2016) observe a fusion of cognitive and affective aspects of trust in the literature. They find studies on process based, characteristic based, institutional based, calculus based, knowledge based, identification based, affective, cognitive, history based, category based and role rule based trust. From the discussion of the many different perspectives on trust, it follows that the model drawn below is only one of the many possible views on trust. It serves as a heuristic for our own research by integrating elements that play a role in the trust process. This designation and the arrangement of the elements follows a decision that could also be made differently without being wrong. The following assumptions (Blöbaum 2016) determine the model (Fig. 6): – Trust is understood as a relationship between trustor and trustee. – Trust is understood as a process and therefore has a time dimension. – Trust is situational. You can trust the trustee in one situation with regard to a certain aspect, but not necessarily in another situation with regard to a different aspect. – Trust is based on the voluntary nature of trustor and trustee. – Trust is linked to risk in two respects: it is influenced by the trustor’s risk propensity and the act of trust is risky in the sense that the expectations associated with it on the part of the trustor may not be realized. – A trustor is always an individual with free will. Trustees can be persons, groups, organizations, institutions, systems. The starting point of a trust process (Fig. 2), in which a relationship between trustor and trustee is established, is an occasion, a motive, which results in the necessity to take the risk of trust. The specific situation is embedded in a specific context. The context includes the setting: Which trustees are potentially available?

16

B. Blöbaum

Fig. 2 Embedding of trust

Which communicative situation is present? Which channels are available? Face-toface? Media mediated? Digital? Presence? Distance? etc. The situation and context also include the significance that entering into a relationship of trust has for the trustor: “The degree to which you trust a particular person to do a particular thing will vary inversely with the degree to which you must rely, for the motivation or justification of your trusting response, on reasons that concern the importance, value or necessity of having such a response” (Hieronymi 2008, p. 213). Among the personality traits that can influence the trust process on the part of the trustor (Fig. 3), a distinction must be made between general and more specific characteristics such as sociodemographic factors (age, gender, education, place of residence, income, etc.) and the experience-related characteristic of socialization. Finally, the personality traits of the propensity to trust and the willingness to take risks are influenced by personal experience as well as by the specific situation. A further set of factors on the part of the trustor can be described as situationrelevant knowledge. Situation-relevant knowledge is not necessarily linked to education (which is part of the personality traits) but to such knowledge that is part of the trust decision in the specific situation. An extensive specific knowledge may keep the risk low. Low knowledge of the reference point is likely to require high trust. Experience has to be distinguished from knowledge. Experiences here do not refer to socialization experiences but to those that are directly connected with the object of trust. If there is no experience with a trustee, be it a person, an organization or a system, this does not exclude a relationship of trust. In such situations there is a higher risk. In practice, it is probably more likely that such a specific case of trust will be related to similar situations, somewhat by comparison or extrapolation. In terms of experience, a distinction can be made between primary and secondary experiences. Primary experiences are those from previous contacts with the trusted partner. Secondary experiences are those made second-hand: one knows someone who has

Some Thoughts on the Nature of Trust: Concept, Models and Theory

17

Fig. 3 Trust features of the trustor

experience with the trustee, one forms an opinion based on evaluations on the Internet, etc. Like experience, the last bundle of factors in Fig. 3 is also related to the object of trust: the perception and evaluation of trustworthiness. These antecedents, which can become experiences, are very diverse. They can range from a person’s appearance (think of racial profiling) to clothing (think of the white coats worn by doctors) to the location of a property (think of some of the scoring used in lending). Such characteristics are understood here as social and cultural factors in the perception of trustworthiness. The factors of perceived ability, benevolence and integrity, already compiled by Mayer et al. (1995), also appear to be suitable and proven in this

18

B. Blöbaum

Fig. 4 Trust features of the trustee

context; although they were compiled for trust-based relationships in organizations (companies), they also have a high general explanatory power. The trust process is always unidirectional from the trustor to the trustee. Even in close romantic or personal relationships, where there is mutual trust, the trust is always directed from one partner to the other: parents trust their children, children trust their parents. The reasons for trust in such relationships may be quite different and are based on different characteristics. The focus of the social sciences is on relationships of trust, moving from a trustor to a trustee as an object of trust. Regarding the trustee (Fig. 4), those features by which trustworthiness can be determined are of particular interest in trust research. A first distinction must be made between the objects of reference: individuals, organizations, systems.

Some Thoughts on the Nature of Trust: Concept, Models and Theory

19

Fig. 5 Process of trust

Sociodemographic factors are more likely to be representable in the case of individuals (e.g. via the CVs of role holders in organizations). In the case of organizations, institutions, systems and other collective structures, organization- and systemspecific characteristics such as history, standards, ethics, compliance, etc. will become the focus of the analysis. As already explained for the trustor, ability, benevolence and integrity are trustrelevant factors for the presentation of trustworthiness on the trustee's side. This introduced terminology is nuanced somewhat differently for the model presented here: as competence, which describes the ability and expertise of the trustee; as reliability, which means consistency in the trustee’s performance; and as professionalism, which means specific competence in a field of action in combination with ethical principles beyond egocentric interests. The demonstration of competence, reliability and professionalism in the performance of services will signal trustworthiness for individuals as well as for systems and organizations. Abilities and integrity can be cognitively condensed to a certain extent by symbols and reputation-enhancing decorations: awards, rating stars, prizes, seals of approval, certificates, etc. These are condensed representations of trust-enhancing elements that should be interpreted by the potential trust-giver as positive signals for the granting of trust. What occurs between trustor and trustee is the process of trust (Fig. 5). This process consists of four stages: prior to, but still a condition for, the trust relationship is an expectation, intention or goal that can only be achieved by granting trust. The trust process itself is initiated by the decision to trust, followed by the specific act of trust. The separation of intention, decision and action is primarily analytical in nature. In the practice of trust, it will hardly be possible to separate these steps. But only when the decision to trust is transformed into the act of trust does a risk arise, making the trustor vulnerable. An evaluation concludes the trust process. A positive evaluation means that the expectation associated with the initiation of the process has been realized. This experience is incorporated into the evaluation of future actions. If the evaluation is negative, if the expectation placed in trust has been disappointed, this will lead to greater caution or distrust in similar situations. Trust is thus linked to a practice, to social decision-making and action. It is therefore not reduced to an attitude. A combination of the components results in the following model heuristic (Fig. 6).

Fig. 6 Trust: relationship and process

20 B. Blöbaum

Some Thoughts on the Nature of Trust: Concept, Models and Theory

21

The model illustrates some of the factors influencing the trust relationship. It does not tell us anything about the empirical correlations. A whole range of assumptions, hypotheses or research questions can be formulated on the basis of existing knowledge and plausible assumptions. Is the propensity to trust a stable personality factor or is it differently pronounced depending on the situation? What is the relationship between the willingness to take risks and the propensity to trust? Does age influence the propensity to trust and the willingness to take risks? This is only intended to indicate how multifaceted and multidimensional research on trust relationships can be. The model mentions elements and factors that play a role in trust relations. However, it does not indicate how these operations are carried out and with what effects. In this respect, the model provides at best an abstract basis as a preliminary stage to empirical validation. As shown above, trust gains in clarity when it has a specific point of reference. The same applies to models. A general model of trust is much more abstract than a domain-specific trust model. This can be shown with a model of trust in online journalism. The antecedents of trustworthiness in online journalism have been developed from journalism research, especially from the discussion on journalistic quality and how procedures in journalism are influenced by digitization (Grosser et al. 2016). These factors, which are related to the programmes of the journalistic system are linked in the model to general elements of trust, such as personality traits of the trustor, situational aspects and risk. The factors of perceived trustworthiness in the journalistic system on the part of the public have the character of hypotheses in this model that can be tested empirically (Fig. 7). The claim of empirical testability is also linked to another model that Wintterlin (2017, 2019) and Wintterlin and Blöbaum (2016) developed to analyse the trust of journalists in their (social media) sources. “At the centre of the model, stands the perceived likelihood of positive or negative outcomes the trustor gains, from weighing factors in the relationship with the trustee (that is, the perceived trustworthiness) and factors outside the relationship that makes it uncertain (that is, the perceived risk)” (Wintterlin and Blöbaum 2016, p. 83). In this model, which relates to the journalistic work process, more precisely to research in journalism, risk is located at two points: as an influencing factor in considering consequences before making a journalistic decision based on previous experience with a journalistic source, and as a cognitive decision to enter into a relationship of trust. Wintterlin calls this “reflexive trust or trust intention” (Wintterlin and Blöbaum 2016, p. 83). On the part of the journalist as a trustor, test criteria for source credibility known from journalism research are supplemented: predictability, similarity, affiliation to the community, as well as further content indicators (Wintterlin and Blöbaum 2016, p. 84) (Fig. 8). The model emphasizes the reflexive character of trust and, as a specification, has—like the other models sketched above—similarities with the trust model of Mayer et al. (1995). The empirical studies of Wintterlin (2019) on journalistic trust in Internet sources confirm the relationships shown in the model. (Other models for

Fig. 7 Model of trust in online journalism. Source: Grosser et al. 2016, p. 65

22 B. Blöbaum

Fig. 8 Model of reflexive trust. Source: Wintterlin and Blöbaum 2016, p. 84

Some Thoughts on the Nature of Trust: Concept, Models and Theory 23

24

B. Blöbaum

the study of trust can be found in this book, for example, in the contributions of Janzik and Quandt, Raue et al., Busch et al. and Querfurth et al.) The presentation of different models on trust shows in summary: – Models arrange and structure the object area. – Trust models vary between abstract and more or less specific representations of the factors and relationships, depending on the area of application. – They are more descriptive than explanatory. The scope of the outlined models corresponds in a certain way to their degree of abstraction. While the general model on trust as a relationship construct generally refers to situational and contextual factors, the model on trust in online journalism specifies trust relationships in a particular media sector. Even more specific is the scope of the model on trust in journalistic sources (especially social media sources). Here, one (of many) journalistic working actions is the subject of trust research. For all their differences, the models have features in common, which also result from the fact that they come from a scientific family: They design trust as a process and relationship, they link trust with risk and assign experiences and perceptions as well as personality traits a fixed place in the decision to trust.

4 Trust: A Theory? In the light of the dominance of empirical research, theoretical work has a difficult position in the scientific community. This is shown in many disciplines by the fact that there are few journals that are dedicated to theoretical discussions, while the journals that prefer empirical contributions clearly predominate. If one uses “model” and “theory” as interchangeable terms, as Jaccard and Jacoby (2010, p. 29) do, then the model developed above would also be a theory of trust. However, difficulties in model building have already been pointed out: the factors described are probably not completely exhaustive, their relative influence is unclear, they are rather descriptive and quite complex. When discussing the theory of trust, the question arises as to its aim: A general theory of trust? Many specific theories of trust—for certain situations, for various contexts, for individuals, organizations, systems? For trust as a process? For trust as a state? Or domain-specific trust for politics, science, sports, etc.? Further questions related to theory would be: What is the relationship between (established) disciplinary theories, for example, from psychology, sociology, economics, communication science and trust theory? Is trust part of a theory of society? Is it an aspect of social or organizational psychological theories? In theory building, a distinction should be made between perspectives. With regard to journalism, for example, the following question should be asked: Does recipients’ trust in journalism (e.g. Grosser et al. 2016) differ from the trust that journalists have in sources (e.g. Wintterlin 2019, 2020)? Or that journalists have in their colleagues? Should there be a theory on trust in media? About trust in a newspaper? In an editorial

Some Thoughts on the Nature of Trust: Concept, Models and Theory

25

office? In an editor? In a reporting field like politics or sports? In the content of individual articles? Theories usually focus on phenomena that cannot be perceived through the senses. In this respect, trust would be a candidate for a theory. A theory stands for a “linguistic entity that orders the phenomena of a subject area in a propositional or conceptual form and describes the essential properties of the objects belonging to it and their relationships to one another” (Mittelstraß 2018, p. 20). If one uses this benchmark, then the model would not meet the theory requirements, because there is neither consensus on what the essential features of trust are nor on the relationships between them. Theories are tools in the process of scientific research. Range, reference points, elements, degree of abstraction, etc.: What dimensions a theory should have is a scientific decision. The same applies to the question of empirical validity. Here a view is preferred according to which relationships between elements constituting a relationship of trust can be empirically tested. Trust and the factors that create trust (but also prevent trust) thus become empirically variables that can be validated. In this way, influences that promote trust can then be distinguished from those that prevent trusting relationships. This then makes it possible to distinguish between high and low trust. The problems mentioned above in the discussion of the concept of trust on the one hand illustrate the volatility of empirical trust research. On the other hand, they also show that respondents show differentiated attitudes in their self-assessments, for example, when they rate their trust in science higher than their trust in politics or the media. Working on a theory of trust does not mean constructing a coherent model. Rather, it is about designing a heuristic that bundles influencing factors and puts them into order. The usefulness of theories and models for the research process is an important touchstone. “Theoretical expressions are valued to the extent that they serve as useful guides to the world we experience” (Jaccard and Jacoby 2010, p. 31). The authors condense from the relevant literature three test criteria, among others: “First, internally, the theory must be logically consistent; that is, the theoretical statements within the conceptual system must not be contradictory, nor must the theory lead to incompatible predictions. Second, the theory must be in agreement with the known data and facts. Third, the theory must be testable; that is, a theory must ultimately be subject to empirical evaluation” (Jaccard and Jacoby 2010, p. 32). Even after several decades, there is still no certainty in research regarding a theory of trust. The micro-perspective, which focuses on the individual (dominant in psychology), the meso-perspective, which focuses on organizations such as companies and groups (dominant in economics) and the macro-perspective, which focuses on systems (in sociology), together with findings from numerous other disciplines such as sports science, computer science, history and communication science have led to a multitude of elements and fragments on the way to a general theory of trust (Endress 2002; Hartmann and Offe 2001; Luhmann 1988; Sztompka 1999). Contributions to the synthesis and assessment of trust often come to the conclusion that the perspectives on the subject are very heterogeneous. Perhaps it is precisely this multiperspectivity on trust that prevents successful theory production.

26

B. Blöbaum

The trust model outlined above is intended to identify basic concepts and factors that influence the trust process—even if the specific relationships between the factors are excluded in detail. It has a didactic function. The description cannot explain trust—but it can help to understand the trust process. For theories and models on trust, the following applies: in principle, pluralism is possible here and even promotes knowledge. From a corpus of theories of the same rank, a suitable one (or more) can be chosen depending on the cognitive interest, taking into account the respective explanatory power. The usefulness of models and theories must also be demonstrated in trust research in scientific practice.

5 Implications Trust as an object of investigation and topic has had a remarkable career in science and media coverage in recent years. Regardless of the motives behind individual studies and media reports: If trust is a phenomenon with which an increasingly complex and interconnected, globalized and fast-moving society reacts to these developments, the expansion of trust research in science and trust reporting in the media can be interpreted as a reaction to perceived threats. As concerns about social cohesion grow, the experience of social community may decline, established institutions in nations are questioned with regard to their ability to solve problems and questions of trust come up. Trust is one of the endangered resources in the process of social change. It is therefore proclaimed as a desirable outcome in relations with role holders, institutions and systems. Therefore, trust is closely linked to perceptions of crisis. In many publications the semantics of trust remain unquestioned. The discussion of surveys on trust suggests that this is the case: not everything labelled with trust actually contains trust. What is labelled as trust in surveys—even in our own research—is often perhaps more like agreement, confidence, an opinion on something. If by trust we understand a processual social relationship that involves a risky action, then trust gains a specific and situational significance—and loses its status as a passe-partout. For research purposes, it can be concluded from this that the broad and diverse concept of trust needs to be explained in detail. The connection of trust to a specific context and situation should be given greater consideration. This also includes naming the reference point of trust and the purpose of trust. Furthermore, it would be beneficial to supplement the predominantly quantitatively oriented research with qualitative survey methods. In-depth interviews, group discussions, focus group interviews and the method of thinking aloud can help to explore what individuals understand by trust and whether and what risks they perceive. Models can help to structure the research process and highlight those factors that are considered relevant to the issue at hand. It remains to be seen whether this helps to obtain a more precise picture of trust. Hartmann (2001, p. 24) is rather sceptical: “It seems to be much easier to first define what trust is in a relatively abstract way than to prove that it is precisely the elements contained in the definition that impose

Some Thoughts on the Nature of Trust: Concept, Models and Theory

27

themselves, for example, on an analysis of the conditions under which trust is created”. If building trust is to be understood as a process, then research must focus more on how to include the temporal dimension. The selective measurements of empirical trust research have so far hardly reflected this process. The operationalization of trust in research remains a challenge.

References Adam, B., Beck, U., Van Loon, J., & Van Loon, B. (2000). The risk society and beyond: Critical issues for social theory. London: Sage. Bachmann, R., & Zaheer, A. (2013). Handbook of advances in trust research. Cheltenham: Edward Elgar Publishing. Barbalet, J. (2019). The experience of trust: Its content and basis. In M. Sasaki (Ed.), Trust in contemporary society (pp. 11–30). Leiden: Brill. Bauer, P. C., & Freitag, M. (2018). Measuring trust. In E. M. Uslaner (Ed.), The Oxford handbook of social and political trust (pp. 1–28). Oxford: Oxford University Press. Blöbaum, B. (2016). Key factors in the process of trust. On the analysis of trust under digital conditions. In B. Blöbaum (Ed.), Trust and communication in a digitized world. Models and concepts of trust research (pp. 3–25). Cham: Springer. Blöbaum, B. (2018). Bezugspunkte von Medienvertrauen. Media Perspektiven, 12, 601–607. Delhey, J., Newton, K., & Welzel, C. (2011). How general is trust in “most people”? Solving the radius of trust problem. American Sociological Review, 76(5), 786–807. Dietz, G., & den Hartog, D. N. (2006). Measuring trust inside organisations. Personal Review, 35, 560. Endress, M. (2002). Vertrauen. Bielefeld: Transcript Verlag. Engelke, K. M., Hase, V., & Wintterlin, F. (2019). On measuring trust and distrust in journalism: Reflection of the status quo and suggestions for the road ahead. Journal of Trust Research, 9(1), 66–86. Erikson, E. H. (1993). Childhood and society. New York: WW Norton & Company. Faus, R., Mannewitz, T., Storks, S., Unzicker, K., Vollmann, E. (2019). Schwindendes Vertrauen in Politik und Parteien. Eine Gefahr für den gesellschaftlichen Zusammenhalt. Gütersloh Fisher, C. (2018). What is meant by ‘trust’ in news media? In Trust in media and journalism (pp. 19–38). Wiesbaden: Springer VS. Frevert, U. (2013). Vertrauensfragen: Eine Obsession der Moderne. Munick: CH Beck. Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. Giddens, A. (2013). The consequences of modernity. New York: John Wiley & Sons. Giere, R. N. (1999). Using models to represent reality. In L. Magnani, N. Nersessian, & P. Thagard (Eds.), Model-based reasoning in scientific discovery (pp. 41–57). Dordrecht: Kluwer. Grosser, K. M., Hase, V., & Blöbaum, B. (2016). Trust in online journalism. In Trust and communication in a digitized world (pp. 53–73). Cham: Springer. Hartmann, M. (2001). Einleitung. In M. Hartmann & C. Offe (Eds.), Vertrauen: die Grundlage des sozialen Zusammenhalts (pp. 7–34). Herausgeber: Campus Verlag. Hartmann, M. (2020). Vertrauen. Die unsichtbare Macht. Frankfurt am Main: S. Fischer. Hartmann, M., & Offe, C. (2001). Vertrauen: die Grundlage des sozialen Zusammenhalts. Herausgeber: Campus Verlag. Hieronymi, P. (2008). The reasons of trust. Australasian Journal of Philosophy, 86(2), 213–236. Hughes, R. I. G. (1997). Models and representation. Philosophy of Science, 64, 325–336. IfK Trendstudie. (2018, 2019). Unveröffentlichte Forschungsergebnisse des Instituts für Kommunikationswissenschaft. Münster: Westfälische Wilhelms-Universität Münster.

28

B. Blöbaum

Jaccard, J., & Jacoby, J. (2010). Theory construction and model-building skills: A practical guide for social scientists. New York: Guilford Publications. Luhmann, N. (1979). Trust and power. New York: Wiley. Luhmann, N. (1988). Familiarity, confidence, trust: Problems and alternatives. In D. Gambetta (Ed.), Trust. Making and breaking cooperative relations (pp. 94–107). Oxford: Basil Blackwell. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. Meyen, M. (2020). Die Erfindung der Glaubwürdigkeit. Umfragen zur Medienbewertung in Deutschland seit 1945. In A. Blome, T. Eberwein, & S. Averbeck-Lietz (Eds.), Medienvertrauen. Historische und aktuelle Perspektiven (pp. 59–75). Berlin: De Gruyter. Miller, A. S., & Mitamura, T. (2003). Are surveys on trust trustworthy? Social Psychology Quarterly, 66, 62–70. Mittelstraß, J. (Ed.). (2016). Enzyklopädie Philosophie und Wissenschaftstheorie: Bd. 5: Log–N. New York: Springer. Mittelstraß, J. (Ed.). (2018). Enzyklopädie Philosophie und Wissenschaftstheorie: Bd. 8: Th-Z. New York: Springer. Möllering, G. (2006). Trust: Reason, routine, reflexivity. Bingley: Emerald Group Publishing. Morgan, M. S., & Morrison, M. (1999). Perspectives on natural and social science. Cambridge: Cambridge University Press. PytlikZillig, L. M., & Kimbrough, C. D. (2016). Consensus on conceptualizations and definitions of trust: Are we there yet? In E. Shockley, T. M. Neal, L. M. PytlikZillig, & B. H. Bornstein (Eds.), Interdisciplinary perspectives on trust: Towards theoretical and methodological integration (pp. 17–47). Cham: Springer. Saam, N. J., & Gautschi, T. (2015). Modellbildung in den Sozialwissenschaften. In N. Braun & N. J. Saam (Eds.), Handbuch Modellbildung und Simulation in den Sozialwissenschaften (pp. 15–60). Wiesbaden: Springer VS. Sasaki, M. (2019). Trust in contemporary society. Leiden: Brill. Schoorman, F. D., Mayer, R. C., & Davis, J. H. (2007). An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32(2), 344–354. Schoorman, F. D., Wood, M. M., & Breuer, C. (2015). Would trust by any other name smell as sweet? Reflections on the meanings and uses of trust across disciplines and context. In B. Bornstein & A. Tomkins (Eds.), Motivating cooperation and compliance with authority (pp. 13–35). Springer International Publishing. Simmel, G. (1978). The philosophy of money. London: Routledge. Simmel, G. (2008). Sociology: Inquiries into the construction of social forms (Vol. 2). Leiden: Brill. Sztompka, P. (1999). Trust: A sociological theory. Cambridge: Cambridge University Press. Uslaner, E. M. (2018). The study of trust. In E. M. Uslaner (Ed.), The Oxford handbook of social and political trust (pp. 1–13). Oxford: Oxford University Press. Warren, M. (2018). Trust and democracy. In E. M. Uslaner (Ed.), The Oxford handbook of social and political trust (pp. 75–94). Oxford: Oxford University Press. Wintterlin, F. (2017). Trust in distant sources: An analytical model capturing antecedents of risk and trustworthiness as perceived by journalists. Journalism. https://doi.org/10.1177/ 1464884917716000. Wintterlin, F. (2019). Quelle: Internet: Journalistisches Vertrauen bei der Recherche in sozialen Medien. Baden-Baden: Nomos Verlag. Wintterlin, F. (2020). Trust in distant sources: An analytical model capturing antecedents of risk and trustworthiness as perceived by journalists. Journalism, 21(1), 130–145. Wintterlin, F., & Blöbaum, B. (2016). Examining journalist’s trust in sources: An analytical model capturing a key problem in journalism. In B. Blöbaum (Ed.), Trust and communication in a digitized world (pp. 75–90). Cham: Springer.

Methodological and Practical Challenges of Interdisciplinary Trust Research Friederike Hendriks, Bettina Distel, Katherine M. Engelke, Daniel Westmattelmann, and Florian Wintterlin

Abstract Trust plays a pivotal role in many different contexts and thus has been investigated by researchers in a variety of disciplines. In this chapter, we provide a comprehensive overview of methodological approaches to investigating trust and its antecedents. We explain how quantitative methods may be used to measure expectations about a trustee or instances of communication about trust efficiently, and we explain how using qualitative measures may be beneficial to researching trust in less explored contexts and for further theory development. We further point out that mixed methods research (uniting both quantitative and qualitative approaches) may be able to grasp the full complexity of trust. Finally, we introduce how agent-based modeling may be used to simulate and predict complex trust relationships on different levels of analysis. We elaborate on challenges and advantages of all these different methodological approaches to researching trust and conclude with recommendations to guide trust researchers in their planning of future investigations on both situational trust and long-term developments of trust in different contexts, and we emphasize why we believe that such undertakings will benefit from interdisciplinary approaches. Keywords Trust · Measurement of trust · Quantitative research · Qualitative research · Mixed methods research · Agent-based modeling F. Hendriks (*) Department of Chemistry Education, IPN – Leibniz Institute for Science and Mathematics Education, Kiel, Germany e-mail: [email protected] B. Distel Department of Information Systems, University of Münster, Münster, Germany e-mail: [email protected] K. M. Engelke · F. Wintterlin Department of Communication, University of Münster, Münster, Germany e-mail: [email protected]; fl[email protected] D. Westmattelmann Center for Management, University of Münster, Münster, Germany e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 B. Blöbaum (ed.), Trust and Communication, https://doi.org/10.1007/978-3-030-72945-5_2

29

30

F. Hendriks et al.

1 Introduction Trust plays a role in many different fields, and it is often unavoidable. One of the most cited definitions of trust states that “trust is a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behaviors of another” (Rousseau et al. 1998, p. 395). The extraordinary role of trust in our modern and digitized societies can be exemplified with the global pandemic caused by the virus SARS-CoV-2 in 2020 (that occurred while this chapter was being written). Citizens must not only trust that their governments will decide on effective measures in order stop the virus from spreading, that scientists will produce reliable and informative research about the virus and measures to contain it, and that journalists will adequately inform them about both political decision making and scientific findings. With the recommendation that people should work from home where possible, team leaders have to trust that co-workers reliably pursue their jobs, while workers in turn must trust that team leaders would advise them well on the new (mostly digital) workflow. Besides digital technology for team collaboration being used, cell phone apps for keeping track of the spread of the virus were developed, resulting in debates about the apps’ trustworthiness and data security. What do these different examples of trust have in common, and how could trust be captured by measurement in each of these cases? The examples above share central features of trust (even though there is no definition of trust yet that is accepted by researchers of all disciplines, see Rompf 2015). Instances in which trust is required are characterized by vulnerability or risk, in spite of which the trustor relies on the trustee or surrenders control, which is only possible by the trustor forming positive expectations about the trustee – the object of trust (Rompf 2015). However, the examples also show that trust may be directed at different “objects.” Following micro, meso, and macro levels of analysis, trust can be directed at people, people as role holders, organizations, institutions, or social systems (Blöbaum 2016; Endreß 2002). First, trust may be directed at people one knows or communicates with in person, in our example, co-workers. Second, especially in digital settings, one often must trust people as role holders, in our example politicians, scientists, and journalists who report on the issue. Third, objects of trust could be organizations (e.g., the trustor’s employing company) or institutions (e.g., health care, scientific institutions). Fourth, the trustor might also trust that reliable systems are in place to control for risks, i.e., that the government makes well-informed decisions, and that other citizens abide by the new-established rules. In this chapter, we also argue for adding technology to the list of trust objects (see Sect. 2.1.3). For example, a person may place trust in technology when she downloads and uses the app (Oldeweme et al., 2021). Finally, trust in various objects may also be the topic of conversations or communication (such as journalistic news coverage). In the following sections, we will briefly summarize methods of measuring and researching trust in all these different instances. In doing so, we will refer to the interdisciplinary background of the Research Training Group “Trust and Communication in a Digitized World,” which collects researchers from Communication

Methodological and Practical Challenges of Interdisciplinary Trust Research

31

Science, Psychology, Information Systems, Business Economics, and Sports Sciences1. In Sect. 2, we summarize quantitative approaches for measuring trust. Besides introducing psychometric approaches for measuring trust in people (as role holders) and technology, we briefly refer to implicit and behavioral measurement. Also, we describe the quantitative measurement of mediated trust in journalistic news coverage via content analysis in Sect. 2.3. In Sect. 3, we describe qualitative approaches to measuring trust, e.g., gathering data via interviews and analyzing data using content analyses, followed by approaches that mix qualitative and quantitative measurement (Sect. 4). In Sect. 5, we report how computer simulations and agent-based modeling (ABM) in particular contribute to trust research. Finally, advantages and challenges—and resulting implications—for the interdisciplinary conceptualization and measurement of trust are summarized in Sect. 6.

2 Quantitative Measurement of Trust: Advantages and Challenges Quantitative measurement has developed as a result of the philosophical notion that concepts in the social sciences and psychology could be assessed in a systematic manner, namely using scientific methods (Kline 1998), and thus are measurable via numerical quantification. First, in psychology, attitudes have long been measured using psychometric assessment. Items are used which require subjects to give ratings on a scale. Several items comprise an instrument, because a certain number of items is required to assess the complexity of a concept and possibly identify sub-scales (we provide an example of a psychometric measurement in Sect. 2.1.2). Using psychometric instruments to survey a large sample of individuals allows researchers to estimate the extent of the manifestation of a concept in a population and to compare between samples by statistical testing (e.g., how much German vs. British people trust the government), or to other psychological concepts (e.g., how people’s trust in the government relates to political literacy). Further, it allows for experimental testing of variables that may influence trust. Possible (significant) differences between or within groups may then be quantified as effect sizes. Overall, this permits verification of results. Second, quantitative measurement also allows systematic examination of communication. Quantitative content analysis may be used to assess collected data in such a way that instances of communication are assigned numerical values, which allows for statistical testing in similar ways as those that were described for psychometric measurement. For example, one could assess the quantity of certain trust frames in news coverage (for further elaboration, see Sect. 2.3) or in social media

1

Other overviews of the measurement of and research on trust (in different disciplinary contexts) can be found elsewhere (e.g. trust and distrust in journalism: Engelke et al. 2019; trust in organizational settings: Lyon et al. 2015; McEvily and Tortoriello 2011).

32

F. Hendriks et al.

entries. In some disciplines or contexts, quantitative content analyses are also used for systematic analysis of data collected by researchers in interviews or focus groups (similarities and differences between quantitative and qualitative content analysis will be briefly elaborated on in Sect. 3). Quantitative measurement enables researchers to repeat and thus replicate other’s studies, and it allows for meta-analysis to estimate the size of an effect over a number of investigations. Further, quantitative measures can be tested for reliability (the consistency of the measure) and validity (the correspondence of the measure with the real-word concepts it aims to grasp). However, while psychometric measurement enables the researcher to make comparisons across time and individuals, or across instances of communication, some level of inter-individual difference might not be uncovered (for example, an individual understanding of the term “trust,” see Sects. 3 and 4). Scientific uncertainty due to (among other reasons) lack of reliability and validity of measurement will always go along with empirical research (see Walker 1990). This is one limitation that can only be reduced, not eliminated.

2.1

Psychometric Measurement

In the following section, we review the quantitative measurement of trust, by referring to the model by Mayer et al. (1995), which highly influenced later conceptualizations of trust, e.g., Rousseau et al. 1998). They defined trust as the willingness of the trustor to be vulnerable to another entity (the trustee), based on the trustor’s own propensity to trust (Sect. 2.1.1) and the positive expectations about the trustworthiness of the trustee (Sects. 2.1.2 and 2.1.3). After briefly explaining the implicit measurement of trust (Sect. 2.1.4), we give examples on measuring trust as behavior (Sect. 2.2).

2.1.1

Characteristics of the Trustor

Besides socio-demographics, most researchers consider trust propensity and risk propensity as important antecedents of trust that pertain to the trustor (Blöbaum 2016). A trustor’s propensity to trust (implying a person’s general willingness to trust others) might influence both the intention to trust in and expectations about the trustee (Mayer et al. 1995). The measurement of a trustor’s trust propensity has largely relied on the work of Rotter (1967), but improved measures have since been created (e.g., Ashleigh et al. 2012; Frazier et al. 2013). Risk propensity is considered “an individual’s general willingness to take risk” (Das and Teng 2004, p. 108), and there are few instruments for its measurement (for one example see, Meertens and Lion 2008). Depending on the disciplinary context, further characteristics of the trustor may be considered relevant for establishing trust in another entity.

Methodological and Practical Challenges of Interdisciplinary Trust Research

2.1.2

33

Trust in People (as Role Holders)

In the literature of research on trust, trustworthiness of the trustee is likely the most studied concept, but with mixed terminology (e.g., speaker credibility, trustworthiness beliefs, trusting intentions, and trusting behaviors). Preceding (and largely inspiring) the trustworthiness definition by Mayer et al. (1995), there has been a longstanding research tradition asking what makes speakers trustworthy, and thus convincing (“speaker credibility,” e.g., Hovland et al. 1953; for a review of scales, see Pornpitakpan 2004). Based on an extensive literature review taking into account these and other scales measuring speaker credibility, trustworthiness, and related concepts, Mayer et al. (1995) define ability, benevolence, and integrity (ABI) as dimensions of the perceived trustworthiness of a trustee. Ability comprises the trustee’s knowledge and skills in a specific area, integrity the trustee adhering to a set of acceptable standards or principles, and benevolence the goodwill of the trustee and an absence of egocentric profit motives. Mayer and Davis (1999) developed their own questionnaire, in which five to six items for each of the ABI dimensions have to be rated on a 5-point Likert scale (from 1 “Disagree strongly” to 5 “Agree strongly”), which was later re-validated for other domains of interest as well (e.g., for the domain of sports; Dreiskämper et al. 2016). A number of questionnaires have since been developed measuring perceptions of trustworthiness in organizational settings, for example, the trustworthiness of leaders or team members (for a review, see McEvily and Tortoriello 2011). While previous scales—both on speaker credibility and trustworthiness—were mostly optimized for face-to-face situations in which a trustor can directly assess the trustworthiness of speakers, Hendriks et al. (2015) developed an instrument applicable to digital settings in the context of experts’ communication of science, extending previous definitions and operationalizations of trust to a conceptualization of epistemic trust (Hendriks et al. 2016a). Epistemic trust directly follows laypeople’s bounded understanding of science (Bromme and Goldman 2014), which makes deference to expert knowledge necessary. In order to decide whether to believe the claims experts make, assessing the experts’ trustworthiness is a feasible and rational strategy to minimize the risk of being misinformed (Bromme and Gierth in press). The Muenster Epistemic Trustworthiness Inventory (METI; Hendriks et al. 2015) measures laypeople’s ascription of epistemic trustworthiness to an expert and (similar to ABI, see Mayer et al. 1995) comprises the three dimensions expertise (e.g., competent–incompetent), benevolence (e.g., responsible–irresponsible), and integrity (e.g., honest–dishonest). It uses semantic differential-type scales (Döring and Bortz 2016) with antonymous adjectives, which are presented on each side of a scale (from 1 to 7). The three-dimensional factor structure of the METI was developed with exploratory and confirmatory factor analysis, and adequate differentiation between factors was shown (Hendriks et al. 2015). The inventory since has been re-evaluated in further studies, where model fit and reliability replicated well (e.g., Merk and Rosman 2019), and has been used to investigate several questions in the

34

F. Hendriks et al.

context of online science communication. For example, the METI has been employed to investigate whether laypeople adapt their trustworthiness ratings about scientist blog authors who disclose ethical aspects related to their field of study (Hendriks et al. 2016b), who use technical language (Thon and Jucks 2017; Zimmermann and Jucks 2018), who use aggressive language (König and Jucks 2019), or who are attacked in social media (Gierth and Bromme 2020).

2.1.3

Trust in (the Context of) Technology

As digitization has increasingly entered into almost all domains of life, the relevance of trust in technology has also increased considerably. Jarvenpaa et al. (1998) were among the first to introduce the concept of trust to the technological context. First studies investigated interpersonal trust relationships in contexts determined by the use of technology, such as virtual teams (Jarvenpaa et al. 1998) or trust in online vendors (Gefen et al. 2003). Here, the measurement of trust corresponds to the interpersonal trust measurement described in the previous sections. In the technology context, not only individuals and organizations can serve as trustees, but also the technology itself or technological artifacts such as e-commerce websites (Ponte et al. 2015; Wei et al. 2009), e-government systems (Distel 2020), robo-advisory services in the financial sector (Bruckes et al. 2019), or security technology (Jalali et al. 2020). These dimensions differ in terms of the characteristics of the trustee. While interpersonal trust is directed towards a “moral and volitional agent” (McKnight et al. 2011, p. 5), trust in a specific technology is placed into a human-made artifact with limited functionality and without will, respectively, moral agency (see for an overview on trust in technology, see Öksüz et al. (2016)). Based on this fundamental distinction, McKnight et al. (2011) concluded that trust in technology “reflects beliefs that a specific technology has the attributes necessary to perform as expected in a given situation in which negative consequences are possible” (McKnight et al. 2011, p. 7). Beyond that, dimensions of trusting beliefs (i.e., trustworthiness) differ in trusting a technology from trusting people (Mayer et al. 1995; McKnight et al. 2011). As trusting beliefs in interpersonal trust involve moral capability—which technologies lack—Mayer et al.’s (1995) ABI dimensions fall short in this context. As such, trusting beliefs in a specific technology are based on the three characteristics: Functionality, helpfulness, and reliability (McKnight et al. 2011). Although trust in technology and interpersonal trust comprise different dimensions, they are related to each other. In online environments, people usually lack direct interaction with people (i.e., vendors) and thus need to base their decision whether to trust a vendor or not on their beliefs about the technology (Li et al. 2012). Conversely, there can also be transfer effects when an established vendor introduces a new technology. For example, it has been shown that trust in a bank as an organization has a positive effect on the initial trust in the offered robo-advisory service (Bruckes et al. 2019). When conducting studies in the technology context, it should therefore be considered whether interpersonal, respectively, interorganizational trust or trust in technology is relevant or whether even the different

Methodological and Practical Challenges of Interdisciplinary Trust Research

35

dimensions of trust should be considered simultaneously in order to account for transfer effects.

2.1.4

Measurement of Trust as an Implicit Attitude

A trustor’s ascriptions of trustworthiness to a trustee might not only be grounded on extensive reasoning, but may also form heuristically (Kruglanski et al. 2005). Hence, there have been approaches to measure trust similar to the measurement of implicit attitudes or stereotypes (for an example, see, Burns and Conchie 2012). The theoretical idea behind implicit measurement of trust is that attitudinal information is stored in close association with a related object. Therefore, reaction times (e.g., of category sorting) should be facilitated when the trust object in question is congruent with the attitude (e.g., trust) but be impeded when it is not (e.g., distrust). Source credibility has been shown to influence both explicit and implicit evaluations of a product, however, the latter counterintuitively only under high elaboration conditions (Smith et al. 2013). As such—and taking into consideration that implicit measurement may have low reliability (Burns and Conchie 2012)—it is unclear how much additional information is gained with the implicit measurement of trust.

2.2

Measurement of Trust as Behavior

One aspect of the theoretical framework that Mayer et al. (1995) provided has been neglected as of yet, namely the behavioral manifestation of trust (originally “risk taking in a relationship”). As providing a full list of possible trust actions is probably impossible, we provide a few illustrative examples in the following paragraphs. In organizational settings, trust can be measured as cooperative behavior (Lewicki et al. 2006), for example, in a “Trust Game” (Berg et al. 1995). In such a game, one player is given money, which she can decide to share with a second player (fully or partially). In the event that she sends any amount of money (she can also decide against sharing) to another player, this amount is tripled by the experimenter. However, the other player is not obliged to return any sum of money of the amount received. That is, if the first player decides to send money, this behavior includes risk and thus indicates trust behavior. In developmental psychology, epistemic trust has been investigated with children as young as three years of age (for a review, see Harris et al. 2018). In the selective trust paradigm two speakers (e.g., confederates or puppets) provide information. In some experiments, speaker characteristics are varied by providing the child with information about a speaker (e.g., “is a liar”), in others, speakers behave a certain way. For example, one speaker may use wrong names for familiar objects several times (e.g., saying “spoon,” when a cup is presented), while the other speaker names all objects correctly (the knowledgeable informant). When unfamiliar objects are subsequently presented, they are labeled differently by the two speakers. Selective

36

F. Hendriks et al.

trust following inferences about speaker expertise is implied when a child uses the knowledgeable informant’s label when it is asked to name the unfamiliar object. Focusing on trust in information, Westphal and Blöbaum (2016) have argued that trust as a behavior may be an exception in digital settings; measuring it might thus be a difficult endeavor. However, there are instances of risk taking in online settings based on positive expectations about others. For example, trust in virtual interlocutors could be indicated when the trustor opts to self-disclose personal information (Jucks et al. 2016; Moll et al. 2014; Thon and Jucks 2014). Investigating this and similar behavior involved in situations that imply risk taking might be a promising way to further investigate trust in digital settings.

2.3

Measurement of Trust as Depicted in Journalistic News Coverage

Trust relationships are not always based on direct experience, but rather on intermediaries: By depicting trust in their journalistic news coverage, the media as one of the most important intermediaries in society can raise public awareness of trust and thus contribute to its emergence (Blöbaum 2014). Understanding how trust is transported through journalistic news coverage requires knowledge on how trust is depicted, which can be investigated via quantitative content analysis (see Sect. 3 for qualitative content analysis). Quantitative content analysis is a central method within communication research but can be applied in other disciplines as well (Coe and Scacco 2017; Lacy et al. 2015). In this method, features of communication—most often of texts—are categorized systematically by assigning them numerical codes with the purpose of conducting statistical analyses to describe trends and patterns regarding the features (Coe and Scacco 2017; Lacy et al. 2015). This categorization—referred to as the process of coding—is guided by a codebook and can be conducted manually (i.e., by humans) or automatically (i.e., by computers), with each involving its own advantages and disadvantages and combinations of both approaches being possible as well (Coe and Scacco 2017; Lacy et al. 2015; Scharkow 2017; Zamith and Lewis 2015). One concept that is drawn upon to capture the depiction of trust in news coverage is that of public trust (Bentele 1994; Bentele 2008; Bentele and Seidenglanz 2008), which is understood as “the attribution of different degrees of trust or distrust in publicly visible individuals, organizations, thus in actors and social systems” (Bentele and Seidenglanz 2008, p. 56) as the objects of trust. The media—respectively, their news coverage—are one possible trust mediator in that they publicly communicate trust factors in the form of specific characteristics of the objects as well as discrepancies in the objects’ communications or actions (Bentele 1994; Bentele 2008; Bentele and Seidenglanz 2008). Studies have drawn upon this concept and used manual quantitative content analysis to investigate discrepancies depicted in

Methodological and Practical Challenges of Interdisciplinary Trust Research

37

news coverage of DAX 30 companies (Seiffert et al. 2011) or both trust factors and discrepancies dealt with in news coverage of the scandals surrounding Christian Wulff, then President of the Federal Republic of Germany (Grünberg et al. 2015). Another possible way to examine the depiction of trust in news coverage via manual quantitative content analysis is to draw upon the more recently developed concept of trust frames, which are the focus of the remainder of this section2 (Engelke 2018). Journalistic framing is the process of giving orientation to recipients on an issue by selecting and highlighting specific aspects of that issue in news content, the product of which is called a news frame (de Vreese 2005; Entman 1993; Entman 2003). The concept of trust frames builds upon the work of Entman (1993, 2003), who defines four frame elements in which the specific aspects are highlighted: (1) The problem definition, which contains the central aspect that is highlighted— i.e., the problem3—as well as relevant actors; (2) the causal identification, which identifies one or more causes for the central aspect; (3) the evaluation of both the causal actor and the central aspect, and (4) the treatment recommendation, which encompasses both suggested and already taken measures addressing the central aspect and also predicts the effects of these two types of measures. Together, these four elements create a structure within a news article that—if repeated across several news texts—is understood as a news frame (Matthes 2007; Matthes and Kohring 2008). According to Entman (1993, 2003), not all four but at least two elements must be present to constitute a frame. While Entman does not specify which elements are needed, their description illustrates that the problem definition and causal identification are crucial: Without the problem definition and its central aspect, a frame would not provide orientation, and without the causal identification for the central aspect, no structure would be present. Moreover, these two elements are the reference points for which problems and actors are evaluated and for which causes measures should address. Trust frames are thus present in news articles when the central aspect of trust is selected and highlighted with regard to an issue. The problem definition features not only trust (see Sect. 2.1 for a definition) but also the trustor and the trustee, which can both appear as individuals, groups of people, or organizations (Fulmer and Gelfand 2012; Rousseau et al. 1998). Moreover, the trustor can also appear as the general public in those cases where no other explicit actor is mentioned, and trust assumedly

2 The concept of trust frames was previously developed for and applied in Engelke (2018). It is part of the larger concept of trust dimension frames, which additionally encompasses distrust frames and trust problem frames and also includes further actors than those discussed here, namely technologies as objects (see also Sect. 2.1.3) and social systems as both subjects and objects in trust, distrust, or trust problem relationships. The following section is therefore a brief and condensed summary of the more extensive and detailed development, description and application of the concept, which can be found in Engelke (2018). 3 Entman (1993, p. 52) states that the problem definition “determine[s] what a causal agent is doing with what costs and benefits”, which demonstrates that the central aspect is not necessarily negative but can also be positive (see also Matthes 2007). While it would therefore be more precise to speak of the “central aspect definition,” we nevertheless use the established term “problem definition.”

38

F. Hendriks et al.

is of wider public relevance. This is based on the fact that journalism fulfills its function for—and journalistic news coverage therefore is addressed to—society as a whole, including the various actors in it (Mast 2018). The causal identification comprises two aspects: The trustor’s propensity to trust as well as antecedents of trustworthiness, which together can lead to trust (Mayer et al. 1995)—i.e., to the central aspect. One or several of the three antecedents—i.e., ability, benevolence, integrity (Mayer et al. 1995)—can be addressed in a news article. The evaluation of the trustee as the causal actor as well as the evaluation of trust itself can be either positive, negative or ambivalent (Matthes and Kohring 2008). The treatment recommendation, finally, encompasses suggested measures to maintain trust and their predicted effects as well as already taken measures to maintain trust and their predicted effects. Maintaining trust can be achieved by enhancing the trustor’s perception of the trustee’s trustworthiness, specifically by demonstrating the three antecedents (Schoorman et al. 2015). The predictions can again be positive, negative, or ambivalent in that they make statements both about the measures’ effects and about whether they should be endorsed or rejected (Entman 1993; Entman 2003; Matthes 2007; Matthes and Kohring 2008). In accordance with the general understanding of news frames, a trust frame can only be identified if at least the first two elements are present and if the ensuing structure is found in more than one news text. In a manual content analysis (see above), the aspects within the four frame elements can be understood as the features that are coded for analysis. Since there have been repeated calls for more transparency regarding the operationalization of frames (Borah 2011; Matthes 2009), a brief explanation follows on how trust frames are identified. The operationalization of trust frames is divided into two content categories: The frame element category in a first and the trust frame category in a second step. The frame element category contains the variables for all the aspects that can be selected and highlighted within the four frame elements, which are derived deductively from the trust and framing literature. The second content category encompasses two variables: The individual frame and the abstract frame. In the individual frame, the codes for all variables in the frame element category are assembled, resulting in one long numerical code. In the abstract frame, only the codes for the variables of the problem definition and the causal identification as the two crucial frame elements are assembled, resulting in a slightly shorter numerical code. These two variables allow an easy identification and differentiation of repeated structures and of the tendencies within these repeated structures, regardless of their frequency—i.e., which trustors and trustees occur in combination with which antecedents of trustworthiness and (for the individual frames) which evaluations and treatment recommendations are highlighted in combination with the specific problem definition and causal identification. Including the shorter abstract frame allows for more generalized insights, especially when only a small sample of news articles is analyzed and completely repeated structures of all four elements are thus found less often than in larger samples. This approach thus identifies frames through a manual dimension reducing procedure, which is characterized by a manual coding of individual frame elements

Methodological and Practical Challenges of Interdisciplinary Trust Research

39

followed by a reduction of the data to frames, mostly via cluster analysis (Matthes 2007; Matthes and Kohring 2008; see Matthes 2007 as well as Matthes and Kohring 2008, for an overview of further methodological approaches to identify frames). However, because the trust frame concept is designed to be applied with regard to different issues and to identify differences in trust frames as precisely as possible, the approach differs from the procedure suggested by Matthes and Kohring in two ways: (1) The different aspects that can be highlighted in the frame elements were not developed inductively for specific issues (Matthes 2007) but deductively and independently of specific issues by conjoining insights from framing and trust literature and (2) frames are not identified via cluster analysis, which groups together similar but nevertheless slightly different structures (Matthes 2007; Matthes and Kohring 2008), but rather by assemblage. Disclosing trust frames in this manner has several advantages: Coding the aspects within the individual frame elements ensures higher transparency and reliability than coding holistic frames (Matthes 2007; Matthes and Kohring 2008). Moreover, the fact that trust frames can be discerned in different contexts allows comparisons of how trust is used to frame different issues. An application is also possible on news texts from the past to trace the presence of frames over time. Finally, frames have effects on both the individual and societal level (de Vreese 2005; Entman 1993) and discerning how exactly trust is used to frame different issues in the news is therefore a prerequisite to understanding how trust emerges via news as an intermediary. Despite these advantages, the application of the concept is challenging: The operationalization of the elements is very complex due to the different aspects that can be highlighted in each element. Hence, the codebook must be very detailed and provide precise instructions for a consistent application. Additionally, coders must be schooled extensively in order for the content analysis to be reliable, since deviations in only one of the coded aspects automatically lead to a discrepancy in the assembled frame structure. In general, the coding process faces issues common to quantitative content analysis (see Coe and Scacco 2017; Lacy et al. 2015).

3 Qualitative Measures of Trust The role of trust for human reasoning and behavior has not been conclusively determined—neither within nor across disciplines and research topics, which is mainly due to the strong context-sensitivity of the construct “trust” (Bachmann 2015). Whereas many instruments have been developed to quantitatively measure trust in various settings (see above), qualitative methods are less commonly used in trust research (Chang et al. 2016). Qualitative methods, in particular, however, are considered to offer deeper insight into especially context-sensitive topics (Bachmann 2015). This section takes a closer look at selected advantages and caveats associated with the qualitative collection and analysis of data and gives some examples of how qualitative instruments can be used in trust research.

40

3.1

F. Hendriks et al.

Applications of Qualitative Methods

Qualitative methods are used in trust research in various disciplines, such as Psychology (e.g., Breuer et al. 2020; Schiemann et al. 2019), Information Systems (e.g., Distel 2018; Distel et al. 2021 in this volume), Communication Science (e.g., Schwarzenegger 2020), or Public Management (e.g., Vallentin and Thygesen 2017) and they are applied for various purposes. 1. As mentioned, trust is an elusive construct and although researchers come closer to a shared understanding of its meaning, lay people may interpret and use the term in many different ways and meanings incongruent to what researchers actually want to study (e.g., Ashleigh and Nandhakumar 2007). Here, qualitative methods can be helpful as they offer room for uncovering this understanding without imposing any meaning or definition by the researcher (Bachmann 2015). Ashleigh and Nandhakumar (2007), for example, use a qualitative approach to extract different meanings of trust that team members of two organizations ascribe to the term. They derive 13 different constructs that interviewees associated with their trust in teams, between teams, and in technology. However, for each of the three trust settings, the constructs were attributed varying degrees of importance. Almost incidentally, the authors come to the important conclusion that: “Although researchers try to develop questionnaires to measure concepts such as trust, it is considered that the facets of trust are too ambiguous [. . .]” (Ashleigh and Nandhakumar 2007, p. 614). Using qualitative methods in trust research can help to uncover the nuances of human trust that standardized instruments tend to eclipse, because respondents are given the opportunity to use their own words and experiences to describe what trust means to them. 2. Qualitative methods are especially suited to yield in-depth insights into—typically but not exclusively—less explored topics. The setting of qualitative studies is often only roughly pre-structured by the researcher and its course is adapted to the individual situation. In this way, the study participants can largely decide in which context they address trust, how much attention they give to the topic, and what role they ascribe to trust. For example, in a study by Fuchs (2012), the results of expert interviews question the general notion of trust being an important antecedent to e-government adoption. Another study, also conducted in the context of e-government, draws a more nuanced picture of trust in the context of e-government use than prior research (see Distel et al. 2021, in this volume). Not only does it suggest that citizens’ trust in the public administration and in the used technical infrastructure is important, but it also highlights different aspects that make up citizens’ trust in the context of e-government, namely: The overall trust level, trust in the service provider, trust in data security and privacy measures, the provider’s perceived integrity, and perceived risks associated to the use of public (e-)services. 3. Furthermore, qualitative techniques can be used to complement existing research and to increase the validity of existing (quantitative) measures of trust. This can be either done as part of a research project in which qualitative and quantitative

Methodological and Practical Challenges of Interdisciplinary Trust Research

41

methods are combined (see Sect. 4 for further details) or as a research project on its own, deepening already accumulated knowledge. Breuer et al. (2020), for example, use a qualitative approach to uncover antecedents to and consequences of trust in virtual teams as compared to face-to-face teams. Using a qualitative approach, they are able to extend the often-cited model by Mayer et al. (1995) with new constructs. In the context of team collaborations, they find that two additional factors of perceived trustworthiness are important, i.e., transparency and predictability. Furthermore, they uncover specific risk-related behaviors that result from team trust, i.e., disclosure, reliance, and contact-seeking.

3.2

Data Collection

Qualitative data can be collected through various means. The most common method used—not only in trust research—are qualitative interviews (for an overview see, Kvale 2007, Brinkmann 2013). Using a semi-structured or unstructured guideline with open-ended questions revolving around a central topic, the researcher interviews respondents and mimics a natural conversation and, thereby, leaves it up to the interviewee how much attention is given to certain aspects and how much details are provided (see, for example, Distel 2018). For example, Distel et al. (2021, in this volume) asked citizens for their perceptions of public administrations in general and in relation to the provision of e-services. Many of the respondents started to talk about trust in their answers to these questions, highlighting the importance of trust in the context of e-government and use of public e-service. Qualitative interviews can be complemented with further techniques such as thinking aloud (e.g., Schwarzenegger 2020) and reconstruction interviews (e.g., Barnoy and Reich 2020). Breuer et al. (2020), for example, used the critical incident technique to extend the trust-model by Mayer et al. (1995). Critical incidents are situations that are important for the interviewee and have a perceived influence on the respondent’s behavior. During a qualitative interview, interviewees were asked to “[. . .] think of a situation in which [the respondents] trusted or distrusted [. . .] team members” (Breuer et al. 2020, p. 11). Once respondents had visualized this situation, they were asked to report on their team members’ behavior that had impacted their trust or distrust. Depending on the context and aim of a study, qualitative interviews can be combined with many more techniques (see for further examples in trust research Lyon 2015). Another means to gather qualitative data is the use of repertory grids. Ashleigh and Nandhakumar (2007), for example, use this technique to extract the different meanings of trust that team members of two organizations ascribe to the term. The repertory grid technique uses a list of “concrete and discrete representations of the domain one wants to explore, known as elements” (Ashleigh and Meyer 2015). The interviewees are then asked to what extent and why they view two of these elements as similar to each other but different from a third element, yielding constructs underlying these elements (Bachmann 2015). Using this technique, it

42

F. Hendriks et al.

becomes possible to extract meanings of words such as “trust” that are so far not covered by research. When applying qualitative methods to collect data, researchers have to consider a range of aspects in the design of their studies. With regard to trust research, three aspects stand out: Access to critical groups (1), protection of data (2), and subjectivity of data (3). 1. Depending on the context, trust becomes a rather sensitive topic. Thus, researchers might find it difficult to gain access to critical groups of respondents and, once access is gained, to retrieve trust-related information on particularly sensitive topics (Lyon 2015). Lyon et al. (2015) vividly describes how difficult access to critical groups can be, especially when cultural barriers have to be overcome. Researchers have to carefully think about who they want to approach as respondents and how the respondents are approached. 2. Another factor coming into play here is one of data protection. Qualitative studies are built on the insights provided by a few individuals or groups of individuals and sample sizes of 20–30 respondents are not uncommon in some disciplines (Kvale 2007). If a study is conducted within one organization, for example, the researcher has to ensure that by using representative quotes from the study or by describing findings in-depth, other employees or managers are not able to associate these statements with specific individual participants of a qualitative study. 3. Moreover, many qualitative approaches—not only in trust research but in general—rely on subjective evaluations of trust situations (e.g., Distel et al. 2021, in this volume; Breuer et al. 2020). Thus, keeping the interview situation free from imposed meanings and understandings is as important as keeping the analysis and interpretation of material free from imposed meanings. The major asset of qualitative methods can be seen in its focus on the subjective experience which should be reflected in the research results. This subjectivity may, however, create tensions when participants report opposing experiences or perceptions. For example, Distel et al. (2021) report that some of their interviewees trust public administrations and view them as a competent provider of public e-services. Some interviewees, however, state the exact opposite. Thus, researchers have to carefully think about how to handle and interpret these tensions. At the same time, the results have to be summarized and—to some extent— generalized, wherefore data analysis becomes a crucial part in qualitative research on trust.

3.3

Data Analysis

When the subjectivity of data on the one hand and the need for generalizing gathered data on the other hand have to be balanced, content analysis as proposed by Mayring (2015) might be a fruitful approach. Though not specifically developed for trust research, the approach to qualitative content analysis enables the researcher to “let

Methodological and Practical Challenges of Interdisciplinary Trust Research

43

the material speak” and to carve out meanings and interpretations of the study’s subjects. This is particularly important considering the already mentioned elusiveness of the construct “trust” and its variability across contexts (e.g., Ashleigh and Nandhakumar 2007). Similar to quantitative content analysis, the qualitative approach also systematically categorizes features of communication, such as interview transcripts, journalistic material, lyrics, or diary entries. This, however, is done through the assignment of text (or other communication) fragments to (thematic) categories. The focus is less on the distribution of contents across materials but more on the content itself. There are several approaches to qualitative content analysis (see Mayring 2004 for a detailed overview), the most commonly used, however, are inductive and deductive content analysis (see Mayring 2000). Using the deductive approach, the researcher assigns text fragments to theoretically developed categories, whereas inductive content analysis develops these categories from the materials. An example of the inductive approach to content analysis in trust research can be found in Distel et al. (2021). In essence, the material—in this case: Transcriptions of qualitative interviews—is first read by the researcher to get a general idea of the contents. Afterwards, a part of the material is analyzed, and preliminary themes or categories are derived. These themes or categories are then applied to new interview material. This coding round may yield new themes or refinements of the previously identified categories that have to be applied to new and already coded interviews. This iterative process is repeated until all interviews are coded and no further categories or themes arise. The process then results in a category scheme containing overarching themes, definitions of these themes, and exemplary quotes from the material. Along such a category scheme the interpretation of the material is undertaken (see Distel et al. 2021 in this volume for an example). This iterative process can be time-consuming, wherefore computer-assisted qualitative data analysis is more and more common. Using special software such as nVivo or MAXQDA, the researcher can easily categorize and analyze the content. Depending on the software, it also enables mixed-method analyses of originally qualitative data. The application of qualitative methods in trust research is certainly challenging, but also has the potential to deepen our understanding of how trust is built, in which contexts different types of trust are relevant, and which consequences trust between humans, in organizations, and in technology has.

4 Mixed Methods in Trust Research Mixed methods are trending in methodological publications in social sciences. The number of handbooks (e.g., Creswell and Clark 2017; Kukartz 2014; Tashakkori and Teddlie 2010) is increasing rapidly and in 2007, Abbas Tashakkori and Charles Creswell founded the Journal of Mixed Methods Research. In the broadest sense, mixed methods approaches are defined by the combination of quantitative and qualitative research methods (Kukartz 2014).

44

F. Hendriks et al.

Following pragmatist approaches, the application of a method is dependent on how well it is able to answer the research question (Creswell 2003). The use of mixed methods therefore has to be justified by arguing why data from different methods could provide a deeper understanding of a phenomenon. Mixed methods designs try to integrate data from different methods for the purpose of triangulation or complementarity. Triangulation is often used as a substitute for validating findings through different methods (Hammersley 2008); however, the actual strength of combining different methods lies in the complementarity of the findings, which helps to understand different aspects of the research object (e.g., Yauch and Steudel 2003). There are at least three possible ways to make sense of the same phenomenon using different methods: 1. Data can be collected and analyzed in parallel. This approach has the advantage that findings from different methods during the same time period can be used to better understand a phenomenon. According to Tashakkori and Teddlie (2010), concurrent and fully integrated mixed designs can be differentiated with the first one describing an approach where the collection and analysis of the data is performed separately, and the results of the different methods are compiled in the end. The second one tries to interactively combine quantitative and qualitative approaches during the research process. For example, Saunders and Thornhill (2011) used parallel card sort and in-depth interviews to examine trust in organizational change processes and Muethel (2012) describes how the “Board Game Method,” which combines qualitative and quantitative methods (e.g., both definition and ranking of categories), can be used to study trust in different cultural contexts. A special case of fully integrated mixed designs are approaches which transform qualitative into quantitative data for the analysis (for an example using critical incidents to identify the mediating role of ABI in the model of trust repair, see Zachariat 2018). In trust research, these combinations of quantitative and qualitative elements in the process of data collection have the advantage that different types of data from the same time point and from the same respondents can be used to make claims based on multiple datasets. This is especially useful if the researchers study, for example, change processes in organizations where it is crucial to include as many perspectives as possible. 2. Data can be collected first qualitatively and afterwards quantitatively. Findings from the qualitative interviews are used to (a) understand the motivations of the respondents and (b) to specify the standardized measurements. Studies on scale development (Kohring and Matthes, 2007) use this approach as well as studies which look at phenomena where previous research on the topic is scarce (Glanz et al. 2013; Wintterlin 2019). Usually, these studies use different samples in the quantitative and qualitative analysis. In the qualitative analysis, experts or typical representatives of the targeted sample are selected. The quantitative sample is oriented towards representativeness. This approach is especially useful if the researchers examine new phenomena where results of other studies are only partly transferable. For example, if trust researchers want to examine whether

Methodological and Practical Challenges of Interdisciplinary Trust Research

45

trust plays a role at all in a relationship which has not been described with the category of trust before, it is appropriate to use qualitative methods first. 3. Data can be collected first quantitatively and afterwards qualitatively. This approach is used to explain the findings gained from standardized surveys or experiments by using in-depth interviews which uncover the mechanisms behind the effects. In trust research, this approach is especially valid for results which contradict previous expectations or to corroborate the quantitative findings (Romeike et al. 2016). The results of qualitative interviews can also bring more depth to quantitative findings and explain the effects found (for an example of trust in journalism, see Wintterlin et al. (2020)). Mixed methods offer opportunities especially for sensitive research topics such as trust. The combination of qualitative and quantitative methods heightens the possibility to capture the full complexity of trust. It enables researchers to employ a constructivist approach where the perspective of the research object on trust is in the focus. Nevertheless, by the combination with quantitative data, the researcher is also able to make more generalizable claims than with qualitative data alone. However, there are also challenges the researcher has to face which refer to conceptual, methodological, and presentation issues (Teddlie and Tashakkori 2015). Combining quantitative and qualitative methods is often time-consuming and requires forward planning to take advantage of both methods. The presentation of results to editors, conference participants, or colleagues requires more knowledge on the part of the audience and must include a justification of the method with a detailed description of the research process. On the side of the researcher, applying mixed methods is challenging because it requires to be a specialist in both methods and to systematically combine both numeric and text information to answer the research questions (Creswell 2003). Nonetheless, to study such a complex issue as trust, mixed-method designs are very helpful to overcome the shortcomings of using only one method if the researcher assumes that diverse types of data provide a better understanding of the research problem.

5 Modeling Trust: Simulations and Agent-Based Modeling When investigating sensitive issues in social sciences, such as trust in certain contexts, there is often the problem that data on the research object are hardly or only indirectly measurable. Simulation methods offer opportunities to overcome limitations of traditional qualitative and quantitative approaches by creating a realistic depiction of a particular phenomenon. Agent-based modeling (ABM) was developed to cope with the increasing complexity of different problems and is increasingly applied in different disciplines such as the social sciences (Macal and North 2009). When ABM is invoked, a system is established that consists of several actors with autonomous decision-making ability and scope for action, the so-called

46

F. Hendriks et al.

agents (see Wilensky and Rand 2015). Different characteristics can be attributed to the individual agents or whole groups of agents, whereby heterogeneous behaviors can be depicted. On this basis, each agent evaluates his situation and then makes individual decisions at the micro-level. At the macro-level, the interaction of the various agents results in an emergent system behavior that cannot be directly derived from the decision algorithms of the individual agents. By using ABM, it is possible to determine how a system responds to modifications of different conditions. Due to its specific characteristics, ABM allows for a differentiated social science approach, which can be described as “generative,” since both the micro- and macro-levels are addressed (Epstein 1999). Therefore, it is often argued that agent-based modeling “is a third way of science” (Axelrod 1997, p. 21) and could complement traditional deductive (positivism) and inductive (interpretivism) reasoning as methods of discovery (Macal and North 2009). Before developing a simulation model for modeling trust, researchers should first examine whether the decision-making behavior of individuals in the context at hand is suitable for an agent-based analysis. The “guidelines for rigor” formulated by Rand and Rust (2011) offer a valuable approach here. A total of six guidelines have been defined for this purpose, which in turn are divided into “indicative,” “required,” or “sufficient.” “Indicative” means that the benefit of using ABM is increased if the research problem meets these guidelines. “Required,” however, expresses that ABM is not suitable if the research problem does not comply with this guideline. “Sufficient” means that ABM can be considered as one of the few approaches if the research problem meets this guideline. 1. Medium numbers (indicative): There should be as many agents present in the system that no single agent can influence a final result at the macro-level itself. 2. Local and potentially complex interactions (indicative): If local and complex interactions take place within the real overall system, the application of ABM is suitable, for example, to assume experience-dependent actions, the interactions with the surrounding environment resulting from the autonomy of the agents. In doing so, the agents can adapt their properties and behavior based on experience gained. 3. Heterogeneity (indicative): Agent-based models focus on individual agents, which is why each agent can be assigned different characteristics and behaviors. Thus, the groups of agents react differently to the behavior of other actors and to changes in environmental influences, depending on their assignment. 4. Rich environments (indicative): Agent-based models facilitate the representation of a diverse and dynamic environment in which the agents must interact with each other and react to changes. Agents in an agent-based model can be designed to survive in any environment by adapting their behavior. 5. Temporal aspects (required): Since an agent-based model can both model processes and study rapidly changing complex systems, it is necessary to consider temporal aspects and therefore this guideline is required for an agent-based model. In addition, temporal aspects change the decision-making process of an

Methodological and Practical Challenges of Interdisciplinary Trust Research

47

agent, because he or she can incorporate experience and future expectations into the decision. 6. Adaptive agents (sufficient): By using a computer-based ABM, agents are able to adapt not only their behavior during the simulation, but also their strategies in the short term. If an agent makes a different decision than before based on experience and accordingly changes his or her behavior to achieve its goals, it is referred to as an adaptive agent. Since ABM allows researchers to create a complex model of reality, social phenomena can be depicted and analyzed which can only be examined to a limited extent (or not at all) in real world settings using the quantitative and qualitative methods described above (Nooteboom 2015). By applying ABM, illegal behavior such as doping abuse of elite athletes can be modeled realistically and the impact of prize money structures on doping behavior can be determined (Westmattelmann et al. 2020). A range of agent-based models has already been developed to analyze trust (Sutcliffe and Wang 2012). For example, Pahl-Wostl and Ebenhöh (2004) use a heuristic approach in their agent-based model by relating attributes like cooperativeness, fairness, risk aversion, negative and positive reciprocity to trustworthiness. However, most models depict trust in specific contexts, such as supply chains (Kim 2009; Meijer and Verwaart 2005; Tykhonov et al. 2008), electronic markets (Breban and Vassileva 2002; Diekmann and Przepiorka 2005), service consumers (Maximilien and Singh 2005), or peer-to-peer networks (Li et al. 2011). The explanatory power of agent-based systems can be significantly increased by integrating qualitative and quantitative findings from empirical studies (Nooteboom 2015). Thus, qualitative studies can provide a sound framework for deriving the underlying processes in the model. Empirical data from quantitative studies can be integrated to calibrate the model in order to achieve more realistic results (Rand and Rust 2011; for a profound guide to the practical implementation of an agent-based simulation model, see Wilensky and Rand 2015).

6 Measurement of Trust: Recommendations for Future Research This chapter aimed to provide the reader with an overview of different approaches to measuring trust, both in face-to-face and in digital settings (keep in mind our initial examples of trust during a global pandemic). In the following section, we sum up the deliberations by elaborating on some challenges of interdisciplinary trust research. Most importantly, pertinent and appropriate conceptualization of trust should directly inform its measurement. As we have discussed in the different sections in this chapter, there are numerous definitions, conceptualizations, and models of trust (Rompf 2015), and we have provided some examples. Hence, to approach the measurement of trust in any discipline, one must first reflect on the following four

48

F. Hendriks et al.

questions (see also Engelke et al. (2019) for similar conclusions regarding the measurement of trust and distrust in journalism). 1. Is it trust? Central features of trust have already been mentioned above: Trust involves the trustor’s expectations about the trustee, a reliance, dependence, or lack of control, and vulnerability or risk. However, when choosing a measure, trust has to be differentiated from similar concepts, for example, credibility (Engelke et al. 2019; Rieh and Danielson 2007), confidence and risk (Giddens 1990), or the personal stance on an issue (Hendriks et al. 2016a). Similarly, trust and distrust are defined differently, as they are indicated by other criteria and should thus not be measured by using the same scales or relying solely on reversed items (Engelke et al. 2019). Furthermore, for choosing a measurement approach, researchers should consider whether their study is focused on situational trust or the development of trust over an extended time. We discuss this in more detail below. 2. Who or what is the trusted entity? As stated above, trust may be directed at people (as role holders), organizations and institutions, systems or technology. Researchers should be careful not to interchangeably use and even merge trust objects under investigation, such as trust in technology (e.g. websites) and trust in people (e.g., the vendor). Hence, the choice of a measurement should be closely guided by theoretical justification of the definition and level of analysis of the trust object to be studied (see Sect. 2.1.3). In the sections above, we have pointed out trust objects that have often been studied using the described approaches. Researchers might use these examples to guide their own choices of an adequate measurement of trust. 3. Which features of trust do I aim to measure (validity)? By precisely defining both trust and the trusted entity, and choosing the right measurement accordingly, a researcher has taken large steps towards the validity of the assessment in a study. However, trust is a concept that may reflect features of the trustor, assessments of the trustee, and short-term and long-term behavioral choices. In the sections above, we have pointed out adequate measurements for each and a combination of trust features. Further, trust in systems relies on the perception that reliable practices are at work (Blöbaum 2016; Kohring 2004). As such, trust in a system might be guided both by expectations about institutions setting reliable standards and practices, and about people who abide by these. Hence, the measure of trust must reflect the level of analysis at hand. Importantly, when aiming to analyze trust directed at several entities that are not on the same level of analysis (e.g., trust in people in a system and trust in systems), the use of mixed methods approaches (Sect. 4) or ABM (Sect. 5) may be advised, which are able to assess trust relationships on several levels of analysis at the same time. 4. Does the measure of choice allow precise measurement (reliability)? While rules of reliability are easier to set for quantitative measurement (for example, one could attend to internal consistency of a scale or intercoder reliability in quantitative content analysis), in qualitative measurement and simulations, such rules are harder to establish. However, while literature has addressed many ways to test

Methodological and Practical Challenges of Interdisciplinary Trust Research

49

and establish reliability in measurement, we advise researchers to consult earlier research on similar questions; to illustrate: Research on speaker credibility inspired Mayer et al.’s (1995) conceptualization of the ABI-model in an organizational setting. We call on researchers to utilize past research not only in their own, but also in other scientific disciplines. As such, interdisciplinary trust research may be able to tackle the challenges of a trust in digitized settings, if researchers continue to develop their methodological repertoire. Between disciplines, trust objects and levels of analysis will likely vary. Quantitative measures have been developed for measuring trustworthiness in a variety of settings and may be adapted or newly validated for other objects of trust (see Sect. 2.1). Furthermore, quantitative approaches allow for replication of measures in future studies in different fields, as can be nicely illustrated by the example of trust frames. Although the concept of trust frames (see Sect. 2.3) was developed from a Communication Science perspective (Engelke 2018), it can be applied in other disciplines as well: Scholars in the field of Economics, Political Science or Sports Sciences, for example, can use it to discern how trust as the central aspect is used to frame the actions of companies, politicians and parties, or athletes. This, in turn, can help researchers gain a better understanding of how trust emerges in the context of their specific field. Interdisciplinary research in this manner can, of course, pose specific challenges: Since the concept of framing was developed in Psychology and Sociology (Borah 2011) and different understandings of it have emerged in different disciplines, non-communication scholars may have to adapt their own understanding of framing in order to identify trust frames in the manner proposed here. Furthermore, non-communication scholars should familiarize themselves with the use of quantitative content analyses in Communication Science (e.g., Coe and Scacco 2017; Lacy et al. 2015) in order to ensure the proper application of the method. Once these challenges are met, other fields may explore new avenues of research and thus benefit from new insights the concept allows them to gain. Similar to the application of trust frames, in each single study researchers are advised to consider how quantitative, qualitative, and mixed methods approaches as well as simulations will likely have to be adapted to the specifics of the trustor, the trust object, and the context. However, research in any discipline will benefit from learning about research methods used to investigate trust in any other discipline. Because methodological operationalization within a study must follow its central conceptualization of trust, interdisciplinary exchange and transdisciplinary collaborative research will largely benefit the joint theoretical understanding of trust. We close this chapter by pointing out two challenges for trust research that also yield implications for future research: First, research methods often address the trustor’s situational assessments of trust. However, the development of trust over time has been granted less attention. For example, a trustor’s assessment of a person’s trustworthiness is most likely situational, but trust in a system may rather develop over time (Reif 2021). Thus, future research should take into account the trust object when choosing a measurement that allows for changes in trust over time. While measurement of situational assessments of the trustworthiness of people

50

F. Hendriks et al.

(as role holders) is most efficiently done using quantitative measurement (even though measurement at several points in time is certainly possible), the long-term development of trust in a system may be better reflected by content analysis of trust frames in media coverage on the topic of interest (see Sect. 2.3) and—in more bounded settings, such as an organization—may also be examined via qualitative or mixed methods approaches (see Sects. 3 and 4). Therefore, the development of longterm research programs directed at measuring how trust emerges, develops, and even erodes over time—and, as follows, identification of critical situations that hold the potential for changing trust in a trustor–trustee relationship—is much needed (but often hindered by limited resources). Second, further research should address how digitization has impacted both the conceptualization and the measurement of trust. This is another reason for the increasing importance of interdisciplinary trust research. As new actors, institutions, and new ways of communication emerge in the digital sphere, questions of trust— whom to trust and on which grounds—gain importance (Blöbaum 2016). Digital settings are characterized by disembedding of social relationships from time and space (Giddens 1990), which results in a new instance of trust, namely one that may shift and develop over time, and which may then manifest again in some instances. For example, Reif (2021) argues that trust in science (as a system) may benefit from participatory online formats, as it provides instances for immediate connections between single scientists (trust in people) and the trustor. Also, other than in faceto-face interactions where trust between actors may continuously build, in digital settings, trust might be formed and informed in a manifold of ways. Hence, the measurement of trust in a digitized world has to be adapted to the complexity within the occurrence and observability of trust. Thus, future interdisciplinary trust research will have to face the increasing relevance of trust in digitized settings.

References Ashleigh, M. J., Higgs, M., & Dulewicz, V. (2012). A new propensity to trust scale and its relationship with individual well-being: Implications for HRM policies and practices. Human Resource Management Journal, 22(4), 360–376. https://doi.org/10.1111/1748-8583.12007. Ashleigh, M. J., & Meyer, E. (2015). Deepening the understanding of trust: combining repertory grid and narrative to explore the uniqueness of trust. In F. Lyon, G. Möllering, & M. N. K. Saunders (Eds.), Handbook of research methods on trust (pp. 138–148). Cheltenham: Edward Elgar Publishing. https://doi.org/10.4337/9780857932013.00023. Ashleigh, M. J., & Nandhakumar, J. (2007). Trust and technologies: Implications for organizational work practices. Decision Support Systems, 43(2), 607–617. https://doi.org/10.1016/j.dss.2005. 05.018. Axelrod, R. (1997). Advancing the art of simulation in the social sciences. In R. Conte, R. Hegselmann, & P. Terna (Eds.), Simulating social phenomena (pp. 21–40). Berlin: Springer. https://doi.org/10.1007/978-3-662-03366-1_2. Bachmann, R. (2015). Utilising repertory grids in macro-level comparative studies. In F. Lyon, G. Möllering, & M. N. K. Saunders (Eds.), Handbook of research methods on trust

Methodological and Practical Challenges of Interdisciplinary Trust Research

51

(pp. 170–177). Cheltenham: Edward Elgar Publishing. https://doi.org/10.4337/ 9781782547419.00025. Barnoy, A., & Reich, Z. (2020). Trusting others: A pareto distribution of source and message credibility among news reporters. Communication Research. https://doi.org/10.1177/ 0093650220911814. Bentele, G. (1994). Öffentliches Vertrauen - Normative und soziale Grundlage für Public Relations [Public trust - normative and social basis for public relations]. In W. Armbrecht & U. Zabel (Eds.), Normative Aspekte der Public Relations. Grundlegende Fragen und Perspektiven. Eine Einführung [Normative aspects of public relations. Fundamental questions and perspectives. An introduction] (pp. 131–158). Opladen: Westdeutscher Verlag. https://doi.org/10.1007/9783-322-97043-5_7. Bentele, G. (2008). Trust of publics. In W. Donsbach (Ed.), The international encyclopedia of communication. London: Wiley. https://doi.org/10.1002/9781405186407.wbiect061. Bentele, G., & Seidenglanz, R. (2008). Trust and credibility - Prerequisites for communication management. In A. Zerfass, B. van Ruler, & K. Sriramesh (Eds.), Public relations research. European and international perspectives and innovations (pp. 49–62). Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-531-90918-9_4. Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity, and social history. Games and Economic Behavior, 10, 122–142. https://doi.org/10.1006/game.1995.1027. Blöbaum, B. (2014). Trust and journalism in a digital environment. Reuters Institute for the Study of Journalism Working Papers. Retrieved from https://reutersinstitute.politics.ox.ac.uk/sites/ default/files/Trust%20and%20Journalism%20in%20a%20Digital%20Environment_0.pdf Blöbaum, B. (2016). Key factors in the process of trust. On the analysis of trust under digital conditions. In B. Blöbaum (Ed.), Trust and communication in a digitized world. Models and concepts of trust research (pp. 3–25). Cham: Springer. https://doi.org/10.1007/978-3-31928059-2_1. Borah, P. (2011). Conceptual issues in framing theory: A systematic examination of a decade’s literature. Journal of Communication, 61(2), 246–263. https://doi.org/10.1111/j.1460-2466. 2011.01539.x. Breban, S., & Vassileva, J. (2002, July). A coalition formation mechanism based on inter-agent trust relationships. In Association for Computing Machinery (Ed.), Proceedings of the first international joint conference on autonomous agents and multiagent systems: Part 1 (pp. 306–307). New York: ACM Digital Library. https://doi.org/10.1145/544741.544812. Breuer, C., Hüffmeier, J., Hibben, F., & Hertel, G. (2020). Trust in teams: A taxonomy of perceived trustworthiness factors and risk-taking behaviors in face-to-face and virtual teams. Human Relations, 73(1), 3–34. https://doi.org/10.1177/0018726718818721. Brinkmann, S. (2013). Qualitative interviewing. Understanding qualitative research. Oxford: Oxford University Press. Bromme, R., & Gierth, L. (in press). Rationality and the public understanding of science. In M. Knauff & W. Spohn (Eds.), The handbook of rationality. Cambridge, MA: MIT Press. Bromme, R., & Goldman, S. R. (2014). The public’s bounded understanding of science. Educational Psychologist, 49(2), 59–69. https://doi.org/10.1080/00461520.2014.921572. Bruckes, M., Westmattelmann, D., Oldeweme, A., & Schewe, G. (2019). Determinants and barriers of adopting robo-advisory services. In International conference on information systems (ICIS 2019). Munich: AIS eLibrary. Burns, C., & Conchie, S. (2012). Measuring implicit trust and automatic attitude activation. In F. Lyon, G. Möllering, & M. N. K. Saunders (Eds.), Handbook of research methods on trust (pp. 239–248). Cheltenham: Edward Elgar Publishing. https://doi.org/10.4337/ 9780857932013.00032. Chang, J. H.-Y., Yang, H., Yeh, K.-H., & Hsu, S.-C. (2016). Developing trust in close personal relationships: Ethnic Chinese’s experiences. Journal of Trust Research, 6(2), 167–193. https:// doi.org/10.1080/21515581.2016.1207543.

52

F. Hendriks et al.

Coe, K., & Scacco, J. M. (2017). Content analysis, quantitative. In J. Matthes, C. S. Davis, & R. F. Potter (Eds.), The international encyclopedia of communication research methods (pp. 346–356). Hoboken: Wiley Blackwell. https://doi.org/10.1002/9781118901731. iecrm0045. Creswell, J. W. (2003). Research design. Qualitative, quantitative, and mixed methods approaches. Thousand Oaks, CA: Sage Publications. Creswell, J. W., & Clark, V. L. P. (2017). Designing and conducting mixed methods research. Thousand Oaks, CA: Sage Publications. Das, T. K., & Teng, B. S. (2004). The risk-based view of trust: A conceptual framework. Journal of Business and Psychology, 19(1), 85–116. de Vreese, C. H. (2005). News framing: Theory and typology. Information Design Journal, 13(1), 51–62. https://doi.org/10.1075/idjdd.13.1.06vre. Diekmann, A., & Przepiorka, W. (2005, August). The evolution of trust and reputation: Results from simulation experiments. Third ESSA Conference, Koblenz, Germany. Retrieved from https://econwpa.ub.uni-muenchen.de/econ-wp/exp/papers/0508/0508005.pdf Distel, B. (2018). Bringing light into the shadows: A qualitative interview study on citizens’ non-adoption of e-government. Electronic Journal of E-Government, 16(2), 98–105. Distel, B. (2020). Assessing citizens’ non-adoption of public e-services in Germany. Information Polity, 2020, 1–22. https://doi.org/10.3233/ip-190214. Distel, B., Koelmann, H., Schmolke, F., & Becker, J. (2021). The role of trust for users’ adoption of public e-services. In B. Blöbaum (Ed.), Trust and communication in a digitized world: Empirical results and implications (p. XX). Hamburg: Rowohlt. Döring, N., & Bortz, J. (2016). Forschungsmethoden und Evaluation in den Sozial- und Humanwissenschaften [Research methods and evaluation in the social sciences and humanities]. Wiesbaden: Springer. https://doi.org/10.1007/978-3-642-41089-5. Dreiskämper, D., Pöppel, K., & Strauß, B. (2016). Vertrauen ist gut: Entwicklung und Validierung eines Inventars zur Messung von Vertrauenswürdigkeit im Sport [Trust is good: Development and validation of an inventory to measure trustworthiness in sports]. Zeitschrift für Sportpsychologie, 23(1), 1–12. https://doi.org/10.1026/1612-5010/a000156. Endreß, M. (2002). Vertrauen [Trust]. Bielefeld: Transcript Verlag. https://doi.org/10.14361/ 9783839400784. Engelke, K. M. (2018). Die journalistische Darstellung von Vertrauen, Misstrauen und Vertrauensproblemen im Kontext der Digitalisierung. Theoretische Entwicklung und empirische Erfassung von Vertrauensdimensions-Frames [The media’s depiction of trust, distrust, and trust problems within the context of digitalization. Theoretical development and empirical analysis of trust dimension frames]. Baden-Baden: Nomos. https://doi.org/10.5771/ 9783845291857. Engelke, K. M., Hase, V., & Wintterlin, F. (2019). On measuring trust and distrust in journalism: Reflection of the status quo and suggestions for the road ahead. Journal of Trust Research, 9(1), 66–86. https://doi.org/10.1080/21515581.2019.1588741. Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51–58. https://doi.org/10.1111/j.1460-2466.1993.tb01304.x. Entman, R. M. (2003). Cascading activation: Contesting the white house’s frame after 9/11. Political Communication, 20(4), 415–432. https://doi.org/10.1080/10584600390244176. Epstein, J. M. (1999). Agent-based computational models and generative social science. Complexity, 4(5), 41–60. https://doi.org/10.1002/(SICI)1099-0526(199905/06)4:53.0.CO;2-F. Frazier, M. L., Johnson, P. D., & Fainshmidt, S. (2013). Development and validation of a propensity to trust scale. Journal of Trust Research, 3(2), 76–97. https://doi.org/10.1080/ 21515581.2013.820026. Fuchs, G. (2012). Lost youth? Attitudes towards and experiences with e-government: The case of German university students. In M. Gasco (Ed.), 12th european conference on e-government (ECEG 2012) (pp. 251–258). Barcelona: Academic Conferences Ltd..

Methodological and Practical Challenges of Interdisciplinary Trust Research

53

Fulmer, C. A., & Gelfand, M. J. (2012). At what level (and in whom) we trust: Trust across multiple organizational levels. Journal of Management, 38(4), 1167–1230. https://doi.org/10.1177/ 0149206312439327. Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS Quarterly, 27(1), 51–90. https://doi.org/10.2307/30036519. Giddens, A. (1990). The consequences of modernity. Cambridge: Polity Press. Gierth, L., & Bromme, R. (2020). Attacking science on social media: How user comments affect perceived trustworthiness and credibility. Public Understanding of Science, 29(2), 230–247. https://doi.org/10.1177/0963662519889275. Glanz, J. M., Wagner, N. M., Narwaney, K. J., Shoup, J. A., McClure, D. L., McCormick, E. V., & Daley, M. F. (2013). A mixed methods study of parental vaccine decision making and parent-provider trust. Academic Pediatrics, 13(5), 481–488. https://doi.org/10.1016/j.acap. 2013.05.030. Grünberg, P., Muxfeldt, C. H., Eichmann, S., Weber, F., Müller, M., & Wecker, M. (2015). Die Causa Wulff - eine Vertrauensanalyse der Medienberichterstattung und des Social Media Diskurses [The Wulff affair - A trust analysis of the news coverage and social media discourse]. In R. Fröhlich & T. Koch (Eds.), Politik - PR - Persuasion. Strukturen, Funktionen und Wirkungen politischer Öffentlichkeitsarbeit [Politics - PR - Persuasion. Structures, functions and effects of political public relations] (pp. 285–303). Wiesbaden: Springer. https://doi.org/10. 1007/978-3-658-01683-8_14. Hammersley, M. (2008). Troubles with triangulation. In M. M. Bergman (Ed.), Advances in mixed methods research (pp. 22–36). London: Sage. https://doi.org/10.5860/choice.51-2973. Harris, P. L., Koenig, M. A., Corriveau, K. H., & Jaswal, V. K. (2018). Cognitive foundations of learning from testimony. Annual Review of Psychology, 69(1), 251–273. https://doi.org/10. 1146/annurev-psych-122216-011710. Hendriks, F., Kienhues, D., & Bromme, R. (2015). Measuring laypeople’s trust in experts in a digital age: The Muenster Epistemic Trustworthiness Inventory (METI). PLoS ONE, 10(10), 1–20. https://doi.org/10.1371/journal.pone.0139309. Hendriks, F., Kienhues, D., & Bromme, R. (2016a). Trust in science and the science of trust. In B. Blöbaum (Ed.), Trust and communication in a digitized world. Models and concepts of trust research (pp. 239–251). Berlin: Springer. https://doi.org/10.1007/978-3-319-28059-2_8. Hendriks, F., Kienhues, D., & Bromme, R. (2016b). Evoking vigilance: Would you (dis)trust a scientist who discusses ethical implications of research in a science blog? Public Understanding of Science, 25(8), 992–1008. https://doi.org/10.1177/0963662516646048. Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion: Psychological studies of opinion change (Vol. 19). Yale: Yale University Press. https://doi.org/10.2307/ 2087772. Jalali, M. S., Bruckes, M., Westmattelmann, D., & Schewe, G. (2020). Why employees (still) click on phishing links: Investigation in hospitals. Journal of Medical Internet Research, 22(1), e16775. https://doi.org/10.2196/16775. Jarvenpaa, S. L., Knoll, K., & Leidner, D. E. (1998). Is anybody out there? Antecedents of trust in global virtual teams. Journal of Management Information Systems, 14(4), 29–64. https://doi. org/10.1080/07421222.1998.11518185. Jucks, R., Linnemann, G. A., Thon, F. M., & Zimmermann, M. (2016). Trust the words: Insights into the role of language in trust building in a digitalized world. In B. Blöbaum (Ed.), Trust and communication in a digitized world. Models and concepts of trust research (pp. 225–237). Cham: Springer. https://doi.org/10.1007/978-3-319-28059-2_13. Kim, W. S. (2009). Effects of a trust mechanism on complex adaptive supply networks: An agentbased social simulation study. Journal of Artificial Societies and Social Simulation, 12(3), 4. https://doi.org/10.18564/jasss.1756. Kline, P. (1998). The new psychometrics: Science, psychology, and measurement. New York, NY: Routledge. https://doi.org/10.4324/9781315787817.

54

F. Hendriks et al.

Kohring, M. (2004). Vertrauen in Journalismus [Trust in journalism]. Konstanz: UVK Verlagsgesellschaft. Kohring, M., & Matthes, J. (2007). Trust in news media: Development and validation of a multidimensional scale. Communication Research, 34(2), 231–252. https://doi.org/10.1177/ 0093650206298071. König, L., & Jucks, R. (2019). Hot topics in science communication: Aggressive language decreases trustworthiness and credibility in scientific debates. Public Understanding of Science, 28(4), 401–416. https://doi.org/10.1177/0963662519833903. Kruglanski, A. W., Raviv, A. A., Bar-Tal, D., Raviv, A. A., Sharvit, K., Ellis, S., & Mannetti, L. (2005). Says who? Epistemic authority effects in social judgment. Advances in Experimental Social Psychology, 37, 345–392. https://doi.org/10.1016/S0065-2601(05)37006-7. Kukartz, U. (2014). Mixed Methods. Methodologie, Forschungsdesigns und Analyseverfahren [Mixed methods, methodology, research designs, and analytical methods]. Wiesbaden: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-531-93267-5. Kvale, S. (2007). Doing interviews. London: SAGE Publications. https://doi.org/10.4135/ 9781849208963. Lacy, S., Watson, B. R., Riffe, D., & Lovejoy, J. (2015). Issues and best practices in content analysis. Journalism and Mass Communication Quarterly, 92(4), 791–811. https://doi.org/10. 1177/1077699015607338. Lewicki, R. J., Tomlinson, E. C., & Gillespie, N. (2006). Models of interpersonal trust development: Theoretical approaches, empirical evidence, and future directions. Journal of Management, 32(6), 991–1022. https://doi.org/10.1177/0149206306294405. Li, F., Pieńkowski, D., van Moorsel, A., & Smith, C. (2012). A holistic framework for trust in online transactions. International Journal of Management Reviews, 14(1), 85–103. https://doi. org/10.1111/j.1468-2370.2011.00311.x. Li, X., Zhou, F., & Yang, X. (2011). A multi-dimensional trust evaluation model for large-scale P2P computing. Journal of Parallel and Distributed Computing, 71(6), 837–847. https://doi.org/10. 1016/j.jpdc.2011.01.007. Lyon, F. (2015). Access and non-probability sampling in qualitative research on trust. In F. Lyon, G. Möllering, & M. N. K. Saunders (Eds.), Handbook of research methods on trust (pp. 109–117). Cheltenham: Edward Elgar Publishing. https://doi.org/10.4337/ 9780857932013.00017. Lyon, F., Möllering, G., & Saunders, M. N. K. (Eds.). (2015). Handbook of research methods on trust. Cheltenham: Edward Elgar Publishing. https://doi.org/10.4337/9781782547419. Macal, C. M., & North, M. J. (2009). Agent-based modeling and simulation. In Proceedings of the 2009 winter simulation conference (WSC) (pp. 86–98). Austin: IEEE. Mast, C. (2018). ABC des Journalismus [ABC of journalism]. Cologne: Herbert von Halem Verlag. Matthes, J. (2007). Framing-Effekte. Zum Einfluss der Politikberichterstattung auf die Einstellungen der Rezipienten [Framing-effects. The influence of news coverage of politics on recipients’ attitudes]. München: Reinhard Fischer Verlag. Matthes, J. (2009). What’s in a frame? A content analysis of media framing studies in the world’s leading communications journals, 1990-2005. Journalism and Mass Communication Quarterly, 86(2), 349–367. https://doi.org/10.1177/107769900908600206. Matthes, J., & Kohring, M. (2008). The content analysis of media frames: Toward improving reliability and validity. Journal of Communication, 58(2), 258–279. https://doi.org/10.1111/j. 1460-2466.2008.00384.x. Maximilien, E. M., & Singh, M. P. (2005, July). Agent-based trust model involving multiple qualities. In Proceedings of the fourth international joint conference on autonomous agents and multiagent systems (pp. 519–526). Utrecht: ACM Digital Library. https://doi.org/10.1145/ 1082473.1082552. Mayer, R. C., & Davis, J. H. (1999). The effect of the performance appraisal system on trust for management: A field quasi-experiment. Journal of Applied Psychology, 84(1), 123–136. https:// doi.org/10.1037/0021-9010.84.1.123.

Methodological and Practical Challenges of Interdisciplinary Trust Research

55

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. The Academy of Management Review, 20(3), 709–734. https://doi.org/10.5465/AMR.1995. 9508080335. Mayring, P. (2000). Qualitative content analysis. Forum: Qualitative Social Research, 1(2), 20. Mayring, P. (2004). Qualitative content analysis. In U. Flick, E. von Kardorff, & I. Steinke (Eds.), A companion to qualitative research (pp. 266–270). London: Sage. Mayring, P. (2015). Qualitative Inhaltsanalyse: Grundlagen und Techniken [Qualitative content analysis: Foundations and techniques] (12th ed.). Weinheim: Beltz. McEvily, B., & Tortoriello, M. (2011). Measuring trust in organisational research: Review and recommendations. Journal of Trust Research, 1(1), 23–63. https://doi.org/10.1080/21515581. 2011.552424. McKnight, D. H., Carter, M., Thatcher, J. B., & Clay, P. F. (2011). Trust in a specific technology: An investigation of its components and measures. ACM Transactions on Management Information Systems (TMIS), 2(2), 1–25. https://doi.org/10.1145/1985347.1985353. Meertens, R. M., & Lion, R. (2008). Measuring an individual’s tendency to take risks: The risk propensity scale. Journal of Applied Social Psychology, 38(6), 1506–1520. Meijer, S., & Verwaart, T. (2005, June). Feasibility of multi-agent simulation for the trust and tracing game. In International conference on industrial, engineering and other applications of applied intelligent systems (pp. 145–154). Berlin: Springer. https://doi.org/10.1007/11504894_ 22. Merk, S., & Rosman, T. (2019). Smart but evil? Student-teachers’ perception of educational researchers’ epistemic trustworthiness. AERA Open, 5(3), 1–18. https://doi.org/10.1177/ 2332858419868158. Moll, R., Pieschl, S., & Bromme, R. (2014). Trust into collective privacy? The role of subjective theories for self-disclosure in online communication. Societies, 4(4), 770–784. https://doi.org/ 10.3390/soc4040770. Muethel, M. (2012). Mixed method applications in trust research: Simultaneous hybrid data collection in cross-cultural settings using the board game method. In F. Lyon, G. Möllering, & M. Saunders (Eds.), Handbook of research methods on trust (pp. 121–129). Cheltenham: Edward Elgar Publishing. https://doi.org/10.4337/9780857932013.00021. Nooteboom, B. (2015). Agent-based simulation of trust. In F. Lyon, G. Möllering, & M. N. K. Saunders (Eds.), Handbook of research methods on trust (pp. 65–74). Cheltenham: Edward Elgar Publishing. https://doi.org/10.4337/9781782547419.00014. Öksüz, A., Walter, N., Distel, B., Räckers, M., & Becker, J. (2016). Trust in the information systems discipline. In B. Blöbaum (Ed.), Trust and communication in a digitized world. Models and concepts of trust research (pp. 205–223). Cham: Springer. https://doi.org/10.1007/978-3319-28059-2_1. Oldeweme, A., Märtins, J., Westmattelmann, D., & Schewe, G. (2021). The role of transparency, trust, and social influence on uncertainty reduction in times of pandemics: Empirical study on the adoption of COVID-19 tracing apps. Journal of medical Internet research, 23(2),e25893. Pahl-Wostl, C., & Ebenhöh, E. (2004). Heuristics to characterise human behaviour in agent based models. In Proceedings of iEMSs 2004 International Congress: “Complexity and Integrated Resources Management”.Osnabrück, Germany. Retrieved from http://www2.econ.iastate.edu/ tesfatsi/HeuristicsHumanBehaviorABM.PahlWostlEbenhoh2004.pdf Ponte, E. B., Carvajal-Trujillo, E., & Escobar-Rodríguez, T. (2015). Influence of trust and perceived value on the intention to purchase travel online: Integrating the effects of assurance on trust antecedents. Tourism Management, 47, 286–302. https://doi.org/10.1016/j.tourman.2014.10. 009. Pornpitakpan, C. (2004). The persuasiveness of source credibility: A critical review of five decades’ evidence. Journal of Applied Social Psychology, 34(2), 243–281. https://doi.org/10.1111/j. 1559-1816.2004.tb02547.x.

56

F. Hendriks et al.

Rand, W., & Rust, R. T. (2011). Agent-based modeling in marketing: Guidelines for rigor. International Journal of Research in Marketing, 28(3), 181–193. https://doi.org/10.1016/j. ijresmar.2011.04.002. Reif, A. (2021). Mehr Raum für Vertrauen? Potenzielle Veränderungen des Vertrauens in Wissenschaft durch partizipative Onlineumgebungen [More space for trust? Potential changes of trust in science through participatory online environments]. In T. Döbler, C. Pentzold, & C. Katzenbach (Eds.), Räume digitaler Kommunikation [Spaces of digital communication] (pp. 210–243). Köln: Herbert von Halem. Rieh, S. Y., & Danielson, D. R. (2007). Credibility: A multidisciplinary framework. Annual Review of Information Science and Technology, 41, 307–364. https://doi.org/10.1002/aris.2007. 1440410114. Romeike, P. D., Nienaber, A. M., & Schewe, G. (2016). How differences in perceptions of own and team performance impact trust and job satisfaction in virtual teams. Human Performance, 29(4), 291–309. https://doi.org/10.1080/08959285.2016.1165226. Rompf, S. A. (2015). Trust and rationality. Wiesbaden: Springer Fachmedien Wiesbaden. https:// doi.org/10.1007/978-3-658-07327-5. Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35(4), 651–665. https://doi.org/10.1111/j.1467-6494.1967.tb01454.x. Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A crossdiscipline view of trust. Academy of Management Review, 23(3), 393–404. https://doi.org/10. 5465/AMR.1998.926617. Saunders, M. N., & Thornhill, A. (2011). Researching sensitively without sensitizing: Using a card sort in a concurrent mixed method design. International Journal of Multiple Research Approaches, 5(3), 334–350. https://doi.org/10.5172/mra.2011.5.3.334. Scharkow, M. (2017). Content analysis, automatic. In J. Matthes, C. S. Davis, & R. F. Potter (Eds.), The international encyclopedia of communication research methods (pp. 324–338). Hoboken: Wiley Blackwell. https://doi.org/10.1002/9781118901731.iecrm0043. Schiemann, S. J., Mühlberger, C., Schoorman, F. D., & Jonas, E. (2019). Trust me, I am a caring coach: The benefits of establishing trustworthiness during coaching by communicating benevolence. Journal of Trust Research, 9(2), 164–184. https://doi.org/10.1080/21515581.2019. 1650751. Schoorman, F. D., Wood, M. M., & Breuer, C. (2015). Would trust by any other name smell as sweet? Reflections on the meanings and uses of trust across disciplines and context. In B. Bornstein & A. Tomkins (Eds.), Motivating cooperation and compliance with authority (pp. 13–35). New York: Springer International Publishing. https://doi.org/10.1007/978-3-31916151-8_2. Schwarzenegger, C. (2020). Personal epistemologies of the media: Selective criticality, pragmatic trust, and competence–confidence in navigating media repertoires in the digital age. New Media & Society, 22(2), 361–377. https://doi.org/10.1177/1461444819856919. Seiffert, J., Bentele, G., & Mende, L. (2011). An explorative study on discrepancies in communication and action of German companies. Journal of Communication Management, 15(4), 349–367. https://doi.org/10.1108/13632541111183389. Smith, C. T., De Houwer, J., & Nosek, B. A. (2013). Consider the source: Persuasion of implicit evaluations is moderated by source credibility. Personality and Social Psychology Bulletin, 39 (2), 193–205. https://doi.org/10.1177/0146167212472374. Sutcliffe, A., & Wang, D. (2012). Computational modelling of trust and social relationships. Journal of Artificial Societies and Social Simulation, 15(1), 3. https://doi.org/10.18564/jasss. 1912. Tashakkori, A., & Teddlie, C. (2010). SAGE handbook of mixed methods in social & behavioral research. Thousand Oaks: Sage. https://doi.org/10.4135/9781506335193. Teddlie, C., & Tashakkori, A. (2015). Overview of contemporary issues in mixed methods research. In A. Tashakkori & C. Teddlie (Eds.), SAGE handbook of mixed methods in social & behavioral research (pp. 1–42). Thousand Oaks: Sage. https://doi.org/10.4135/9781506335193.n1.

Methodological and Practical Challenges of Interdisciplinary Trust Research

57

Thon, F. M., & Jucks, R. (2014). Regulating privacy in interpersonal online communication: The role of self-disclosure. Studies in Communication Sciences, 14(1), 3–11. https://doi.org/10. 1016/j.scoms.2014.03.012. Thon, F. M., & Jucks, R. (2017). Believing in expertise: How authors’ credentials and language use influence the credibility of online health information. Health Communication, 32(7), 828–836. https://doi.org/10.1080/10410236.2016.1172296. Tykhonov, D., Jonker, C. M., Meijer, S. A., & Verwaart, D. (2008). Agent-based simulation of the trust and tracing game for supply chains and networks. Journal of Artificial Societies and Social Simulation, 11(3), 1–30. Vallentin, S., & Thygesen, N. (2017). Trust and control in public sector reform: Complementarity and beyond. Journal of Trust Research, 7(2), 150–169. https://doi.org/10.1080/21515581.2017. 1354766. Walker, V. R. (1990). The siren songs of science: Toward a taxonomy of scientific uncertainty for decision makers. Connecticut Law Review, 23, 567–627. Wei, T. T., Marthandan, G., Chong, A. Y. L., Ooi, K. B., & Arumugam, S. (2009). What drives Malaysian m-commerce adoption? An empirical analysis. Industrial Management & Data Systems, 109(3), 370–388. https://doi.org/10.1108/02635570910939399. Westmattelmann, D., Sprenger, M., Hokamp, S., & Schewe, G. (2020). Money matters: The impact of prize money on doping behaviour. Sport Management Review, 23(4), 688–703. Westphal, S., & Blöbaum, B. (2016). Trust as an action: About the overrated significance of trust in information sources in a digitized world. In B. Blöbaum (Ed.), Trust and communication in a digitized world. Models and concepts of trust research (pp. 113–124). Cham: Springer. https:// doi.org/10.1007/978-3-319-28059-2_6. Wilensky, U., & Rand, W. (2015). An introduction to agent-based modeling: Modeling natural, social, and engineered complex systems with NetLogo. Cambridge, MA: The MIT Press. Wintterlin, F. (2019). Quelle: Internet. Journalistisches Vertrauen bei der Recherche in sozialen Medien [Source: Internet. Journalists‘ trust when researching in social media]. Baden-Baden: Nomos Verlagsgesellschaft. https://doi.org/10.5771/9783845295121. Wintterlin, F., Engelke, K., & Hase, V. (2020). Can transparency preserve journalism’s trustworthiness? Recipients’ views on transparency about source origin and verification regarding usergenerated content in the news. Studies in Communication and Media, 9(2), 218–240. https://doi. org/10.5771/2192-4007-2020-2-218. Yauch, C. A., & Steudel, H. J. (2003). Complementary use of qualitative and quantitative cultural assessment methods. Organizational Research Methods, 6(4), 465–481. https://doi.org/10. 1177/1094428103257362. Zachariat, S. (2018). Managing trust across levels-empirical propositions for banks [Doctoral dissertation, Universität Münster]. https://d-nb.info/1176629352/34 Zamith, R., & Lewis, S. C. (2015). Content analysis and the algorithmic coder: What computational social science means for traditional modes of media analysis. The Annals of the American Academy of Political and Social Science, 659(1), 307–318. https://doi.org/10.1177/ 0002716215570576. Zimmermann, M., & Jucks, R. (2018). How experts’ use of medical technical jargon in different types of online health forums affects perceived information credibility: Randomized experiment with laypersons. Journal of Medical Internet Research, 20(1), e30. https://doi.org/10.2196/jmir. 8346.

Part II

Trust Research in the Field of Media

Perceptions of Trustworthiness and Risk: How Transparency Can Influence Trust in Journalism Bernadette Uth, Laura Badura, and Bernd Blöbaum

Abstract Trust in the news media and the role of journalism in society is currently highly discussed. Media outlets are concerned about their trustworthiness and consider how to strengthen their audience’s trust. In order to build trust, journalism outlets have two main starting points. We argue that, one, journalism can work on the perceived risk of its performance, two, journalism can try to actively work on its trustworthiness and put out signals that prove it. One influential way to do so is often seen in journalistic transparency. This article examines how transparency influences journalistic trustworthiness, risk, and the formation of trust. In order to shed light on the concepts of trust, risk, and transparency as well as their connection, we conducted a representative survey of 1029 media users in Germany. Our study shows that transparency can be a way for news outlets to increase their trustworthiness and improve risk perceptions, which in turn presents a way to restore and strengthen trust in their work. The results enable us to develop a more defined view of the concepts of risks and trust in journalism, along with indications for journalistic practice on how to work more transparently and to win (back) trust. Keywords Trust · Risk · Transparency · Journalism · Trustworthiness · Representative survey · Risk perception

B. Uth (*) · L. Badura · B. Blöbaum Department of Communication, University of Münster, Münster, Germany e-mail: [email protected]; [email protected]; [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 B. Blöbaum (ed.), Trust and Communication, https://doi.org/10.1007/978-3-030-72945-5_3

61

62

B. Uth et al.

1 Introduction “Who trusts us anymore?”1 This question was discussed by the national German weekly “Die Zeit” in 2015. A couple of years later, the discussion about trust in the news media and the role of journalism in society is still relevant. In today’s increasingly complex society, we face a gap between the knowledge and information needed for everyday life and the amount of information that can be obtained based on available resources (Giddens 1991; Jackob 2012; Warren 1999). By relying on the information conveyed by mass media, citizens are able to solve this problem (Tsfati and Cohen 2005). Journalistic news coverage is an important element in modern society to provide information about relevant issues and thereby helps people to navigate themselves in their environment. However, since recipients cannot check the correctness and authenticity of every bit of information transmitted, they always take on a risk when relying on the information—thus, they need to trust in journalism as a reliable conveyor (Blöbaum 2014; Kohring 2004). In times when trust in journalism is often discussed, media outlets are concerned about their trustworthiness and consider how to strengthen their audience’s trust in them. Journalism outlets that want to work on the trust relationship with their audiences have two main starting points. We argue that, one, journalism can work on the perceived risk of its performance: if recipients hold a better perception of the specific risk connected to the journalistic content, they are in the position to form better, clearer judgments on whether or not to trust the outlet. Two, journalism can try to actively work on its trustworthiness and to put out signals that prove it. One influential way to both increase trustworthiness and sharpen risk perception is often seen in journalistic transparency. It shall provide recipients with a clear view of the various elements that influence the formation of trust in journalism. This article examines which elements constitute the process of trust in journalism and how transparency as a trust-building strategy influences this process—and therefore the trust relationship between journalism and recipients. Based on the empirical results, practical implications for journalism and media outlets will be derived. In the first step, we theoretically define which elements constitute and influence the formation of trust in journalism (Sect. 2.1). In the second step, we take a closer look at transparency as one strategy to work on the trust relationship between journalism and its audience (Sect. 2.2). In the third step, we approach the complex topic of trust in journalism empirically by conducting a representative public survey in Germany that asks about risk, trust, and transparency in journalism (Sect. 3). The results enable us to develop a more defined view of the concepts of risks and trust in journalism, along with indications for journalistic practice on how to work more transparently and to win (back) trust.

1 Author’s translation from the German. https://www.zeit.de/2015/26/journalismus-medienkritikluegenpresse-vertrauen-ukraine-krise.

Perceptions of Trustworthiness and Risk: How Transparency Can Influence. . .

63

Regarding the scope of this article, it is important to note that we define journalism as current and up-to-date informational journalism whose task consists in both the selection and communication of current and relevant issues in order to provide recipients with all the information they need to act in their role as citizens in a democratic society (Esser and Neuberger 2019). This article solely focuses on content produced by professional journalists in organized contexts, meaning that other types of journalism, such as participatory journalism and citizen journalism do not fall within the scope of this study.

2 Elements Constituting and Factors Influencing Trust 2.1

Modeling the Relationship Between Trust and Risk as a Process

Trust is seen as a characteristic of a social relationship between two entities: the trustor, in this case the recipient; and the trustee, in this case the journalistic system (Kohring 2004). Trust in the journalistic system can be focused on three levels: journalistic organizations, specific journalists, and journalistic products and media content (Blöbaum 2014; Uth 2019). The concept of trust cannot be explained without a discussion of the concept of risk (Lewis and Weigert 1985), which is why we suggest to look at the formation of trust as a process that combines trust and risk (Badura and Uth 2019; for other authors conceptualizing processes for trust and/or risk, see, for example, Das and Teng 2004; Fischer 2016; Kee and Knox 1970; McKnight and Chervany 2001). Figure 1 depicts this process. 1. Trust develops in situations in which the trustor needs to make themselves dependent on another individual, organization, or system in order to reach a goal or task that the trustor cannot achieve themself (Rousseau et al. 1998). As the trustor is not able to control or monitor a trustee, they always perceive a risk when relying on the trustee (Mayer et al. 1995). Possible risks associated with trusting journalism include the possibility that the conveyed information might be incomplete, erroneous, or distorted (Badura 2016; Grosser 2016).

Fig. 1 Trust and risk as a process

64

B. Uth et al.

2. The formation of trust is based on the trustor’s expectations of the trustee (Luhmann 1968). These expectations are always focused on the specific action that the trustor expects the trustee to execute and are based on experiences with the trustee in the past (Blöbaum 2016; Hardin 2002; Kohring 2004). In the case of trust in journalism, this means that the trustor expects journalism to perform its systemic task of researching, selecting, and accurately reporting current and relevant topics and facts to the trustor (Grosser 2016). 3. Based on these expectations, the trustor will make two assessments: One, they assess the specific risk associated with the trusting action; two, they will make an assessment of the trustee’s trustworthiness (Mayer et al. 1995). This trustworthiness is assessed via so-called “access points” (Giddens 1991, p. 83): Institutions and systems (such as the journalistic system) are too complex for recipients to directly assess and therefore trust—with access points, the trustor is in touch with specific representatives of a social system. For the journalistic system, access points can be: (1) a journalistic organization, such as a specific news outlet; (2) specific journalists; as well as (3) journalistic products and media content (Blöbaum 2014; Uth 2019).2 Recipients can judge the trustworthiness of these access points by several indicators, antecedents, that serve as cues and signals of trustworthiness (Mayer et al. 1995). Examples for these indicators for each journalistic access point can be: (1) the reputation of the journalistic organization, (2) the ability and expertise of specific journalists, and (3) the quality of the content (Badura et al. 2018; Bogaerts and Carpentier 2013; Jackob 2012; Jakobs 2018). The judgment of trust antecedents is always subjective: while one trustor may judge an entity as very trustworthy, another might feel completely differently and assign low trustworthiness to said entity (Kohring 2004; Sztompka 1999). Furthermore, the assessment of trustworthiness is always situation-specific: while a trustee might be judged as highly trustworthy in one situation, the same trustee might be judged as untrustworthy in another situation that is focused on a different task (Hardin 2002). Parallel to assessing the trustworthiness of journalism, the trustor also assesses the risk connected to the trusting action: This perception is influenced by several factors, such as situational factors (e.g., context, communication situation, familiarity) and the relevance of the trusting action to the trustor (Badura 2016). The higher the perceived personal consequences for the trustor, the higher they (usually) perceive the associated risk (Coleman 1994; Renn 1989). 4. Both the perceived risk and the perceived trustworthiness of the trustee are central to the decision for or against entering into a trust relationship; the two will be weighed against each other in order for the trustor to decide whether they are or are not willing to trust the potential trustee (Mayer et al. 1995, p. 715). In the case of journalism, recipients have to decide whether they are willing to make themselves vulnerable to the system’s function and therefore “to the journalistic 2

Some of these theoretical underpinnings have already been published in Badura et al. (2020).

Perceptions of Trustworthiness and Risk: How Transparency Can Influence. . .

65

system’s selection and communication of current information” (Grosser 2016, p. 1040). Since entering into a trust relationship always implies an acceptance of the associated risk, a willingness to trust can also be seen as a “willingness to take a risk” (Mayer et al. 1995, p. 714). This willingness can be seen as the next step of the trust process and is a central precondition for a trusting action to happen; however, willingness alone is not sufficient for trust, as it does not yet include taking the risk, which is central to trust (Dietz and Den Hartog 2006). “In other words, trust can only matter if it results in specific trusting behaviours that make the trustor vulnerable to the trustee” (Li 2012, p. 102). 5. The risk is only eminent with the concrete trusting action, by which a trustor actually makes themselves vulnerable to the trustee in the form of “an individual’s behavioral reliance on another person under a condition of risk” (Currall and Judge 1995, p. 151). The risk is thus always connected to the actual action of trust, which might not necessarily happen according to the expectations of the trustor (Hardin 2002). The trusting action is also often termed “risk taking in relationship” (Mayer et al. 1995, p. 715). In the case of journalism, the trusting action has to be defined in a rather broad way: it entails a number of processes, like solidifying knowledge, forming attitudes, and even making decisions based on the information provided by journalism (Grosser 2016; Prochazka and Schweiger 2016). The formation of trust is strongly influenced by the third step of the above process, in which the risk associated with trusting the journalistic system as well as the trustworthiness of the journalistic system is assessed.

2.2

Transparency as a Strategy Influencing the Formation of Trust

Several strategies have been discussed in the literature as allegedly helping journalism to clearly communicate and even actively increase its trustworthiness, as well as to enable recipients to make clearer assessments of the risks connected to journalism (see, for example, Blöbaum 2014). One strategy that is often considered to effectively achieve this goal consists of increasing journalistic transparency. Transparency provides recipients with more information on and insights into the journalistic products and the production processes happening inside news outlets (Kovach and Rosenstiel 2001). Transparency should enable recipients to better understand the processes behind the journalistic products; news outlets can provide the audience with further information that enables them to follow the production process and check the information provided themselves (Meier and Reimer 2011). Therefore, implementing transparency should allow recipients to better judge whether their expectations towards journalism and its functions are being fulfilled,

66

B. Uth et al.

as well as put them in a better position to assess the trustworthiness of the journalistic system, its organizations, its roles, and its contents and products (Blöbaum 2014). Several strategies for classifying journalistic transparency have been mentioned in the literature (see, for example, Heikkilä et al. 2012; Karlsson 2010; Meier and Reimer 2011), which we integrate below into one overarching classification of transparency in journalism. We differentiate transparency along the source of transparency (internal transparency, participatory transparency, and external transparency) as well as the access points of the journalistic system to which they are related (organization, journalist, and product).3 First, transparency can be differentiated along who is involved in the process of its creation (see Meier and Reimer 2011): Internal transparency entails a variety of transparency strategies from inside news outlets with which they and their journalists open up about their work and themselves; it is therefore self-referential. This type of transparency can be further differentiated along two dimensions, which Meier and Reimer (2011) term product transparency and process transparency: Product transparency summarizes all types of transparency concerning the final journalistic product, such as naming sources or implementing hyperlinks; process transparency aims at explaining the processes behind creating news and giving behind-the-scenes insights of news outlets. External transparency means that transparency concerning the journalistic system is created by third parties outside of the news outlets, such as external press councils and fact-checking entities (Meier and Reimer 2011). Both internal and external transparency can be implemented in two ways, as Meier and Reimer (2011) emphasize: They can be realized in a unidirectional way— as has been the tradition in journalism for several decades already—such as in the form of source information and hyperlinks. However, due to the new interactivity of the web, transparency can now also be implemented in an interactive and dialogical way that offers recipients the possibility to take part in creating and perceiving transparency. In a similar vein, Karlsson (2010) differentiates between disclosure transparency and participatory transparency: while the former entails news outlets opening up about their work by offering information about it and their products to the audience, the latter includes the audience’s active participation in the journalistic workflows and therefore enables the audience to gather insight into journalistic production processes. Therefore, we include participatory transparency as a third dimension, located between internal transparency and external transparency. The users and their content are not part of the journalistic system; however, their participation can usually be controlled by the news outlets. Examples would be moderated user comments or guided and controlled participation of recipients in the editorial process. In the following analyses, we will distinguish between forms of internal, participatory, and external transparency. These three forms of transparency can be

3

The classification is based on an already-published article from the authors (Badura et al. 2020).

Perceptions of Trustworthiness and Risk: How Transparency Can Influence. . .

67

located on all three earlier mentioned levels of access points to trust in journalism (being news outlets, journalists, and content): 1. On the organizational level, news outlets can create transparency by publishing editorial guidelines and standards that guide their everyday work. This is understood as a unidirectional internal strategy of transparency that does not involve the audience. In order to directly involve the audience, news outlets can open up editorial meetings to the public, either digitally via livestream or by inviting them into the newsroom. External transparency can be created by metajournalistic observations by other news outlets or by watch blogs (for the concept of metajournalistic discourse, see Carlson 2016). 2. On the individual level, transparency can be created by publishing author profiles in which journalists can give insights on their career and personal experiences, in this way demonstrating competence in order to increase trustworthiness. Involving the audience, transparency can be fostered by organizing fora in which the public can ask journalists about their work or accompany them throughout their day. External transparency strategies can, for example, be seen in press councils that watch over journalists’ work and, if necessary, criticize them. 3. On the product level, a long-established way to foster transparency is openly sharing sources, which is facilitated on the web by hyperlinks. Furthermore, supplementary information and documents can be added to journalistic products in order to enable recipients to check the information themselves. One way to actively include the audience and therefore to create participatory transparency on this level is to give the audience the opportunity to participate in the production process, such as by suggesting topics to cover or even by doing research. In terms of external transparency, fact-checkers are one strategy to show additional transparency by verifying the information in journalistic content. Empirical evidence of a connection between implementing transparency elements and trust in journalism remains contradictory: While some studies do find a connection between the use of transparency strategies and trust in journalism, others fail to prove an increase in trust after transparency elements are implemented (for an overview, see Curry and Stroud 2019; Reimer 2017). An experiment by Koliska (2017) found no influence of different transparency strategies on recipients’ trust evaluations at the process or producer level. Karlsson et al. (2014) experimentally tested the effect of both participatory transparency and internal transparency on the product and process levels and were also unable to prove a connection between transparency and the credibility of both sources and messages. However, an experimental study by Meier and Reimer (2011) did show a connection between process transparency and trust, and Curry and Stroud (2019) found that increased transparency (by adding five transparency elements on both the product and process levels) leads to higher credibility evaluations. Chen et al. (2019) also found that transparency at the process level (in the form of information boxes on the journalistic process) positively influences the evaluation and perceived trustworthiness of an article. Furthermore, evidence from qualitative interviews by Wintterlin et al. (2020) shows that for most recipients, transparency about the origin of the source has a

68

B. Uth et al.

positive influence on trustworthiness. However, being transparent about the verification status of an information evokes ambivalent reactions: while some participants consider verification transparency to be a sign of professionalism, in other participants this type of transparency evokes a feeling of skepticism and insecurity, especially in the case of information that could not be verified. Since the connection between transparency and trust in journalism has not yet been conclusively established, this study seeks to contribute to this discussion by further exploring recipients’ views on internal, participatory, and external transparency and trust, with the help of the aforementioned access points in journalism. Since existing research paints a rather diffuse picture of the effects of various transparency strategies, this study uses a new categorization of transparency to exploratively and descriptively assess which transparency strategies are perceived by recipients as relevant: RQ1: Which transparency strategies do recipients rate as important for them? Furthermore, we assume that transparency enables recipients to come to clearer judgments of both journalistic risks and journalistic trustworthiness, which in turn could increase trust. This leads us to the following hypothesis: H1: The perception of transparent journalism is associated with higher trust in the media.

2.3

Trustor Characteristics as a Factor Influencing the Formation of Trust

The whole process of forming trust in journalism depends on the specific trustor; several personal characteristics determine whether and how much someone will trust the journalistic system (see, for example, Jackob 2012). These factors are assumed to involve sociodemographic characteristics of the trustor, such as age, gender, and education; however, research has thus far only found mixed results concerning the influence of these characteristics on trust in journalism (for an overview, see Jackob 2012; Jakobs 2018). Therefore, we propose the following research question: RQ2: Is there a connection between sociodemographic characteristics of the trustor and trust in journalism? Studies have shown that political orientation has a strong influence on trust in journalism in the German media system: recipients on the outer margins of the political spectrum generally show lower trust in the media (Jackob et al. 2019; Reinemann et al. 2017; Schultz et al. 2017). Therefore, we suggest the following hypothesis: H2: Respondents on the margins of the political spectrum will show lower trust in the media. Media usage is often seen as a highly influential factor on trust in the media; since trust is largely based on former interactions and experiences with a trustee, regular

Perceptions of Trustworthiness and Risk: How Transparency Can Influence. . .

69

usage of a certain medium should lead to higher trust in the media (Jackob 2012; Tsfati and Ariely 2014). This leads us to the following hypothesis: H3: Higher usage of a medium will lead to higher trust in the media. Another factor influencing trust in journalism is the general trust propensity: the more one thinks that people are generally trustworthy, the higher one’s trust in the media (Gronke and Cook 2007; Uth and Blöbaum 2019). This leads us to the following hypothesis: H4: General trust propensity will be positively associated with trust in the media. As recipients’ perspectives and characteristics are pivotal in the formation of trust, we decided to address our research questions and hypotheses by conducting a representative survey in Germany. A short outline of the survey is given in the following section.

3 Empirical Evidence 3.1

Methods

In order to answer the research questions and hypotheses, we conducted a public survey with 1029 participants. The data was collected in April 2019. Sampling was done via an online-access panel, and the survey is representative of the German population aged 16–64 years with Internet access (regarding age, gender, education, and region). The sample consists of 51.3% male and 48.7% female participants, with an average age of 40.4 years. The survey focused on up-to-date informational journalism; participants were asked to think of the media they usually consume to stay informed about current events when answering the questions pertaining to their attitude towards the media. The research project on media trust has been collecting data annually since 2017, including data on trust in journalism. Previous data from this survey have already been published (cf. Blöbaum 2018) and presented at conferences. The results presented here refer to the third wave of data.

3.2

Transparency Strategies

One way of enabling recipients to come to clearer judgements of risks connected to journalism and trustworthiness can be seen in journalistic transparency. Therefore, we take a closer look at how recipients perceive the strategies undertaken by journalistic actors to improve risk perception and journalistic trustworthiness assessments and therefore the earlier mentioned process of trust formation (RQ1). Participants were given eight different journalistic transparency strategies and asked to rate how important they perceive each strategy on a 4-point scale (from unimportant to very important, see Fig. 2).

Fig. 2 Assessment of transparency strategies. Reprinted with permission of transcript Verlag. Original source: Badura et al. (2020)

70 B. Uth et al.

Perceptions of Trustworthiness and Risk: How Transparency Can Influence. . .

71

The mean values generally show that all transparency strategies are perceived as rather important (seven out of the eight named strategies receive a mean rating above 2.5 points, which represents the center of the scale). However, what is striking is the relatively high number of participants that could not answer the question: 171 participants chose the “I don’t know”-option for at least one of the eight strategies. This suggests that a large proportion of participants might not have encountered the named strategies thus far. In particular, the concept of the ombudsperson seems to be rather unknown to the participants—16.6% of the participants chose the “I don’t know”-option for this strategy. This may be due to the fact that the institution of the ombudsperson is not yet very widespread in Germany (see Heikkilä et al. 2012). Upon closer examination of the different strategies based on their access points to the journalistic system, it becomes apparent that strategies at the content level are considered to be particularly important: fact-checkers, a form of external transparency, are on average assigned the most relevance (72.1% of all respondents rate factcheckers as rather or very important). Other strategies on the content level are also regarded as very and rather important—such as the possibility to participate in the selection of topics (71.1%) and in the research process (60.9%). Due to the collaboration between journalists and the audience, these strategies can be classified as participatory transparency. In an examination of internal transparency strategies, it becomes clear that the publication of editorial guidelines (an organizational-level strategy) seems to be more important than personal profiles of journalists or blogs in which journalists report about their work (strategies on the level of the individual journalist). The least important transparency strategy is participation in editorial conferences, which would provide profound insights into the editorial processes but would also require a very high degree of active participation by the audience. This suggests that recipients prefer transparency strategies that require less commitment. The mean importance of each strategy—differentiated along access points to the journalistic system as well as along distance to the news outlets—is shown in the figure below (see Fig. 3). In summary, strategies concentrated on the level of the journalistic content are generally the most relevant to the audience, and strategies for product transparency seems to be perceived as more important than those for process transparency. The results also show that recipients are interested in actively participating in the production of the journalistic product, such as by participating in the research process or in selecting topics—forms of participation that might result in additional advantages for editorial offices. What is striking is that the strategy perceived as most important does not involve the news outlets themselves: the strategy of using external fact-checkers. From the recipients’ view, the work of fact-checking entities focuses on the risk of erroneous or distorted information and enables recipients to more accurately assess the risk connected to the use of journalistic information. Recipients seem to value the independent examination service, which is not surprising considering that the accuracy of the information regularly represents one of the

72

B. Uth et al.

Fig. 3 Assessment of transparency strategies, differed along access points and source. Reprinted with permission of transcript Verlag. Original source: Badura et al. (2020)

most important expectations of journalism (Karlsson et al. 2017). Clearly, recipients show less interest in how the product actually comes together and what happens behind the scenes; instead, they prefer being given the chance to minimize the risk of the reception by having additional information and transparency in the final journalistic content.

3.3

The Perception of Transparency and Its Connection with Trust

In addition to asking recipients to assess the personal relevance of several transparency strategies, we seek to explore the connection between transparency and trust in journalism in order to answer H1. Therefore, we asked recipients to assess how transparent they perceive German journalism to be. Since we aim to evaluate which strategies news outlets can actively pursue to build trustworthiness and to improve risk perception, the items focused solely on internal transparency strategies—in other words, we only asked for transparency that can be achieved by the news outlets themselves, and excluded all forms of participatory and external transparency. The operationalization of transparency was organized along the two aforementioned dimensions by Meier and Reimer (2011): product transparency, which involves transparency within and about the journalistic content itself, and process transparency, which involves transparency about the processes behind the products and inside the news outlet. Product and process transparency were operationalized by two statements each (see Fig. 4). To assess product transparency, we asked recipients to state their agreement with whether journalistic products are traceable and transparent, and whether journalists communicate the sources their reporting is based on. To assess

Fig. 4 Perceptions of product and process transparency. Reprinted with permission of transcript Verlag. Original source: Badura et al. (2020)

Perceptions of Trustworthiness and Risk: How Transparency Can Influence. . . 73

74

B. Uth et al.

process transparency, we asked recipients to state their agreement with the statement that journalists communicate how they select their news selection, and whether journalists are transparent about their editorial processes. The results show that the processes (editorial approach and methods, news selection) are perceived as less transparent than the products (journalistic contents as a whole, open communication of sources used). While 34.3 respectively 33.7% of participants do not agree or rather disagree that journalists open up about their production processes or explain how they selected the news they report, only 25.2% of them feel that journalists do not transparently communicate their sources and only 21.4% state that they do not perceive journalistic products as transparent. However, what is striking is the overall ambivalent perception of transparency in journalism: a large share of respondents (35–46.1%) stated that they see transparency as given in some cases, but not in others. In addition, 123 respondents answered one or more of the questions in this block with “I don’t know,” and 42 respondents were even unable to make a statement on any of the four transparency statements, which suggests that some respondents might not yet have had any contact with transparency in journalism and/or might have struggled to recognize transparency elements. This supports the results of an experiment by Grosser et al. (2019), in which some participants experienced difficulty in recognizing the transparent indication of the verification status of information. Furthermore, in qualitative interviews by Wintterlin et al. (2020), some participants stated that they tend to perceive little transparency in everyday life, even if they view transparency as important to them when addressed, as our first query showed. The four items measuring perceptions of product and process transparency are suitable for calculating an index that describes the perceived transparency in journalism from the audience’s perspective, an index that from now on will be called the transparency index (M ¼ 2.944; SD ¼ 0.84; α ¼ 0.869). As already explained, transparent journalism should enable the audience to make clearer judgments of both journalism’s trustworthiness and the associated risks, which is why we expect that perceiving journalism as transparent and trust in the media will be positively associated (see H1). Indeed, we find a positive correlation between the transparency index and trust in the media (r(897) ¼ 0.462; p < 0.001).5 In order to determine whether this connection remains when controlling for other factors that influence trust (see RQ2, H2, H3, and H4), we ran a hierarchical linear regression model with general media trust as the dependent variable (see Table 1). The main independent variable is the transparency index; however, the model also included several other factors that are assumed to have an influence on trust in journalism: sociodemographic characteristics, media usage, political voting

The mean value and the index were calculated excluding the answer option “I don’t know” (n ¼ 934–976). 5 Trust in the media was measured on a five-point scale via the statement: “In general, the media can be trusted.” 4

Perceptions of Trustworthiness and Risk: How Transparency Can Influence. . .

75

Table 1 Regression model DV: Generalized Media Trusta B SE Sociodemographics of the trustor (R2 ¼ 0.004; p ¼ 0.071) Gender 0.167 0.066 Age 0.000 0.003 Political voting intention (R2 ¼ 0.137; p < 0.001)b AfD CDU 0.535 0.121 CSU 0.462 0.183 SPD 0.513 0.124 FDP 0.315 0.138 Bündnis90/Die Grünen 0.409 0.117 Die Linke 0.257 0.128 Others 0.076 0.158 Ineligible to vote 0.315 0.223 Non-voters 0.050 0.155 Media usage (R2 ¼ 0.150; p < 0.001)c Internet news sites 0.008 0.074 News apps 0.077 0.077 Nationwide daily newspapers 0.009 0.076 Local and regional newspapers 0.016 0.081 Weekly newspapers/magazines 0.020 0.082 Radio 0.036 0.080 TV 0.212 0.091 News content on Facebook/Twitter 0.070 0.074 Blogs 0.005 0.093 Trust propensity (R2 ¼ 0.289; p < 0.001)d 0.333 0.034 Perceived characteristics of the trustee (R2 ¼ 0.368; p < 0.001) Transparency index 0.408 0.042

β

P

0.074 0.005

0.011 0.873

(Reference) 0.177 0.082 0.156 0.079 0.140 0.072 0.016 0.047 0.011