Misinformation and Fake News in Education 9781641138512, 9781641138529, 9781641138536

Citation preview

Misinformation and Fake News in Education

A volume in Current Perspectives on Cognition, Learning, and Instruction Daniel H. Robinson and Matthew T. McCrudden, Series Editors

This page intentionally left blank.

Misinformation and Fake News in Education

edited by

Panayiota Kendeou University of Minnesota

Daniel H. Robinson The University of Texas at Arlington

Matthew T. McCrudden Pennsylvania State University

INFORMATION AGE PUBLISHING, INC. Charlotte, NC • www.infoagepub.com

Library of Congress Cataloging-in-Publication Data A CIP record for this book is available from the Library of Congress   http://www.loc.gov ISBN: 978-1-64113-851-2 (Paperback) 978-1-64113-852-9 (Hardcover) 978-1-64113-853-6 (ebook)

Copyright © 2019 Information Age Publishing Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the publisher. Printed in the United States of America

CONTENTS

1 Misinformation and Disinformation in Education: An Introduction..................................................................................... 1 Panayiota Kendeou, Daniel H. Robinson, and Matthew T. McCrudden

SECT I O N I SUSCEPTIBILITY TO MISINFORMATION IN EDUCATION 2 Zombie Concepts in Education: Why They Won’t Die and Why You Can’t Kill Them...................................................................... 7 Gale M. Sinatra and Neil Jacobson 3 Understanding Susceptibility to Educational Inaccuracies: Examining the Likelihood of Adoption Model................................. 29 Alexandra List and Lisa DaVia Rubenstein 4 Psychological Tribes and Processes: Understanding Why and How Misinformation Persists............................................................... 55 Gregory J. Trevors 5 Cognitive Biases in Forensic Science Training and Education......... 81 Candice Bridge and Mark Marić 6 Do Individual Differences in Conspiratorial and Political Leanings Influence the Use of Inaccurate Information?................ 103 David N. Rapp, Megan N. Imundo, and Rebecca M. Adler v

vi   Contents

7 Educational Muckrakers, Watchdogs, and Whistleblowers............. 123 Daniel H. Robinson and Robert A. Bligh 8 Designing Interventions to Combat Misinformation Based on Factors That Increase Susceptibility............................................ 133 Abbey M. Loehr and Andrew C. Butler

SECT I O N I I PRACTICES IN THE SERVICE OF REDUCING MISINFORMATION IN EDUCATION 9 Modeling the Dissemination of Misinformation Through Discourse Dynamics........................................................................... 159 Laura K. Allen, Aaron D. Likens, and Danielle S. McNamara 10 A Nation of Curators: Educating Students to be Critical Consumers and Users of Online Information................................. 187 Jeffrey A. Greene, Brian M. Cartiff, Rebekah F. Duke, and Victor M. Deekens 11 Misinformation in Autism Spectrum Disorder and Education...... 207 Jessica Paynter, Ullrich K. H. Ecker, David Trembath, Rhylee Sulek, and Deb Keen 12 From Theory to Practice: Implications of KReC for Designing Effective Learning Environments..................................................... 229 Jasmine Kim, Reese Butterfuss, Joseph Aubele, and Panayiota Kendeou 13 How Attempting to Reduce Misconceptions in Psychology Reveals the Challenges of Change.................................................... 259 Patricia Kowalski and Annette Taylor 14 Critical Thinking in the Post-Truth Era........................................... 279 Åsa Wikforss 15 Attempting to Reduce Misinformation and Other Inaccuracies in Education....................................................................................... 305 Matthew T. McCrudden About the Editors............................................................................... 319 About the Contributors...................................................................... 321

CHAPTER 1

MISINFORMATION AND DISINFORMATION IN EDUCATION An Introduction Panayiota Kendeou University of Minnesota Daniel H. Robinson The University of Texas at Arlington Matthew T. McCrudden Pennsylvania State University

Today, like no other time in our history, the threat of misinformation and disinformation is at an all-time high. This is also true in the field of education. This book provides recent examples of how misinformation and disinformation manifest in the field of education and offer remedies. To understand the nature of misinformation and disinformation in education, it is important to agree on basic definitions. We draw on recent work Misinformation and Fake News in Education, pages 1–4 Copyright © 2019 by Information Age Publishing All rights of reproduction in any form reserved.

1

2    P. KENDEOU, D. H. ROBINSON, and M. T. McCRUDDEN

in communication sciences, psychology, and education to provide some clarity. Misinformation refers to false information that is shared by a source who has the intent to inform, but is unaware that the information is false or inaccurate (e.g., a false connection, a wrong interpretation, a myth, or a misconception; Waerdle & Derakhshan, 2017). For example, many educators tell students that underlining or highlighting is an effective learning strategy when in fact it has low utility as a learning strategy when used in isolation (Dunlosky et al., 2013). Disinformation refers to false information (e.g., manipulated or fabricated content, misleading information) that is shared by a source who has the intent to deceive and is aware that the information is false (Lazer et al., 2018). For example, some politicians claim that high-stakes testing, combined with teacher and principal accountability systems, will fix K–12 education when in fact there is no evidence to support this practice, and is thus misleading. Fake news, a type of disinformation, refers to “fabricated information that mimics news media content in form but not in organizational process or intent. Fake-news outlets, in turn, lack the news media’s editorial norms and processes for ensuring the accuracy and credibility of information” (Lazer et al., 2018, p. 1094). Unfortunately, the term has also been used by politicians around the world to describe news organizations whose coverage they find disagreeable. These definitions raise an important question; namely, how do misinformation and disinformation manifest in education? To address this question, we find it useful to compare each term along two important dimensions; the accuracy of the information and the intention of the source to either inform or deceive (see Table 1.1). In principle, researchers, educators, and policy makers aim to inform or be informed. Nevertheless, at times, well-intended sources may share information that is inaccurate or incomplete. At other times the desire to influence decisions to adopt certain curricula, assessments, books, instruction, or interventions may result in manipulation, fabrication, and deception. In this volume, several contributions present examples of concepts, ideas, teaching methods, and interventions that despite being “false,” they continue to influence education. An equally important question is “Why are we TABLE 1.1  Misinformation and Disinformation Misinformation

Disinformation

Definition

False information that is shared by a source who has the intent to inform, but is unaware that the information is false or inaccurate.

False information that is shared by a source who has the intent to deceive and is aware that the information is false.

Accuracy of Information

Inaccurate

Inaccurate

Intent of Source

To inform

To deceive or obfuscate

Misinformation and Disinformation in Education    3

susceptible to misinformation?” The contributions in this volume provide in-depth discussions that highlight various sources of susceptibility drawing on social psychology, cognitive science, memory research, motivated reasoning, educational psychology, and communication sciences. Finally, and perhaps the most important question is “What can we do about it?” The contributions in this volume offer various interdisciplinary solutions such as the use of computational linguistics, interventions, audience design, and developing skills such as critical thinking. OVERVIEW OF THE CONTRIBUTIONS In Section I, “Susceptibility to Misinformation in Education,” the collection of chapters focuses on factors that influence the endorsement and persistence of misinformation in dducation. Sinatra and Jacobson identify “zombie concepts” in education to better understand why, despite persistent efforts, such myths continue to enjoy widespread support. List and Rubenstein propose the “likelihood of adoption model” to help understand our susceptibility to educational inaccuracies. Trevors draws on cognitive, motivational, social psychology, and political science literatures to define intentional correction resistance; namely, correction failure that is due to individuals’ intentional rejection of attempted corrections. Bridge and Maric discuss how confirmation bias manifests in forensic science education as a product of the “CSI effect.” Rapp, Imundo, and Adler bring to the forefront conspiratorial ideation and political ideology as individual difference factors that influence susceptibility to misinformation. Robinson and Bligh identify examples of dramatic turnarounds in standardized test scores that turned out to be hoaxes. Finally, Loehr and Butler offer a critical discussion on how the content and characteristics of the misinformation, cognition of the learner, and sociocultural and contextual factors increase susceptibility to misinformation and misconceptions. In Section II, “Practices in the Service of Reducing Misinformation in Education,” the collection of chapters focuses on practices aimed at reducing the impact of misinformation in education. Allen, Likens, and McNamara discuss the promise of using dynamical systems and computational linguistics to model (and possibly combat) the spread of misinformation. Greene, Cartiff, Duke, and Deekens discuss interventions to mitigate misinformation challenges, drawing on research on self-regulated learning, multiple source use, and social-psychological research in education. Kim, Butterfuss, Aubele, and Kendeou use KReC to integrate how text, task, and reader factors can combat misconceptions via audience design. Paynter, Ecker, Trembath, Sulek, and Keen discuss the need for multilevel support to maintain sustained change in terms of reducing or eliminating use of

4    P. KENDEOU, D. H. ROBINSON, and M. T. McCRUDDEN

ineffective or fad practices in the area of autism spectrum disorder. Kowalski and Taylor describe how they apply evidence from research on misconceptions in psychology classrooms and finding practices that work. Wikforss offers an evaluation of critical thinking in the post-truth era and challenges educational systems that encourage skepticism about truth and objectivity in science. Finally, McCrudden offers a critical discussion of the chapters in this section and identifies common themes that emerged in the context of corrective efforts in education. CONCLUDING REMARKS Taken together, we believe that the ideas put forth in this collection of chapters advance our understanding with respect to the current challenges that misinformation and disinformation pose in various education contexts as well as approaches to correction that draw on several literatures. Indeed, researchers have advocated for a multidisciplinary effort to combat the issue of misinformation and disinformation that focuses on both the information ecosystem and individuals as “consumers” of information (Lazer et al., 2018). With respect to the ecosystem, we need to do more to prevent the propagation of misinformation. To do this, we must make systemic changes to our information systems—new safeguards are needed (e.g., filtering algorithms)—that align with current technological advances. Further, as “consumers” of information, we need to develop skills that allow us to effectively separate fact from fiction and, ultimately, support an education system that will train students on critical evaluation skills from a young age. REFERENCES Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning and comprehension: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58. Lazer, D. M., Baum, M., Benkler, Y., Berinsky, A. J., Greenhill, K. M., Menczer, F., . . . Rothschild, D. (2018). The science of fake news, Science, 359(6380), 1094–1096. Wardle, C., & Derakhshan, H. (2017). Information disorder: Toward an interdisciplinary framework for research and policy making (DGI(2017)09). Strasbourg, France: Council of Europe. Retrieved from https://rm.coe.int/information-disorder -toward-an-interdisciplinary-framework-for-researc/168076277c

SECTION I SUSCEPTIBILITY TO MISINFORMATION IN EDUCATION

This page intentionally left blank.

CHAPTER 2

ZOMBIE CONCEPTS IN EDUCATION Why They Won’t Die and Why You Can’t Kill Them Gale M. Sinatra University of Southern California Neil G. Jacobson University of Southern California

The moon landing was faked, the Earth is flat, climate change is a hoax, vaccines cause autism. These are myths you have likely heard before or perhaps endorsed yourself at one time. It is likely that individuals have succumbed to myths for as long as humans have used stories to understand their world (Lévi-Strauss, 1995). Sailors feared sea monsters, ancient people thought eclipses were caused by dragons eating the sun. Prior to Galileo, the Earth was thought to be the center of the universe. Today, with the ubiquity of the Internet, myths propagate quickly and perpetuate indefinitely.

Misinformation and Fake News in Education, pages 7–27 Copyright © 2019 by Information Age Publishing All rights of reproduction in any form reserved.

7

8    G. M. SINATRA and N. G. JACOBSON

Education is not immune to myths. You only use 10% of your brain. You are either a right-brain or left-brain person. Today’s students are “digital natives” and thus are savvy about how to learn with technology. Individuals have their own unique learning styles. These and dozens of other education myths have been well documented (for an overview, see De Bruyckere, Kirschner, & Hulshof, 2015). Misinformation, quackery, and fake news, like myths, are also not new. The famed “snake oil salesman” was a legendary profiteer who aimed to make his fortune off unwitting customers by selling fraudulent goods. Today, with mass media and the Internet, a different type of snake oil is “sold” (either figuratively to persuade or literally to sell books, curriculum materials, or classical music to make infants smarter) and that is misinformation, quackery, and fake news. Sometimes false information is unwittingly shared, but often it is deliberately “weaponized” or spread with malicious intent or for profit. We use the term “zombie concepts” (Kirschner & Neelen, 2018) to refer to education concepts that are notoriously difficult to snuff out of mainstream educational parlance despite both a complete lack of supporting evidence and mounting evidence to the contrary. In this chapter, we address four basic questions about these hard-to-kill ideas: (a) What are zombie concepts?; (b) Why are zombie concepts so appealing?; (c) How do zombie concepts endure?; and finally, (d) Why are they so hard to kill? We conclude with some recommendations for fighting the zombie concept apocalypse. Briefly, to respond to the first question, a zombie concept is a persistent myth that enjoys widespread popular support despite little to no empirical evidence. Examples of zombie concepts include: individuals learn differently depending on their preferred learning style, there are inherent gender differences that account for enrollment differences in STEM, and education kills creativity (see De Bruyckere et al., 2015 for a detailed list of education myths). In regard to the second question, we argue that zombie concepts arise in part because of their tremendous intuitive appeal, comprehensibility, and resonance. Zombie concepts resonate because they offer simple solutions to intractable educational problems, they make intuitive sense, and they offer “explanations” about ourselves (i.e., “Oh, that’s why I’m not so great at math, I’m an auditory learner.”). But, most insidious, they excuse us from addressing persistent educational issues (i.e., Why do some students succeed in school while others struggle?). These intractable issues require difficult and costly solutions (i.e., providing equal access and opportunity for all students to quality instruction that meets their needs) but such time and effort can be avoided by invoking a zombie concept. Next, we turn to these two questions: How do zombie concepts endure? and Why can’t you kill them? Zombie concepts endure because individuals

Zombie Concepts in Education    9

are motivated to have them persist. Perhaps because of their emotional resonance or due to dispositions to seek simple answers, individuals are motivated to keep the concepts alive. Zombie concepts persist in the hearts and minds of individuals due to what Sinatra and colleagues call “the hat trick of change” (Sinatra, Kienhues, & Hofer, 2014). That is, these concepts require individuals to experience conceptual, attitudinal, and epistemic conceptual change. Each type of change is difficult to achieve in its own right, but promoting change in all three is even more challenging. Changing individual minds and attitudes may still be insufficient to kill off zombie concepts. The perpetual motion machine of Internet misinformation also keeps zombie concepts alive. There are websites devoted to these concepts citing plenty of “research” that is of questionable origin as support. Sadly, many teacher education programs teach these concepts as facts, thus passing them on to the next generation of educators. Next, we break these questions down a bit more and delve deeper into our perspective on the answers. WHAT ARE ZOMBIE CONCEPTS IN EDUCATION? Zombie concepts in education are persistent myths that, once animated, perpetuate in formal, nonformal, and informal learning environments with little to no empirical support. De Bruyckere et al. (2015) describe a taxonomy of education zombie concepts including: neuromyths about learning, myths about technology in education, and myths about educational policy. Myths about education include learning styles, boys are naturally better at math than girls, and schools kill creativity. Examples of neuromyths include; we are good multitaskers, we only use 10% of our brain, and babies become smarter if they listen to classical music (Rauscher, Shaw, & Ky, 1993). Myths about technology in education include; the Internet makes us dumber, young people do not read anymore, and nothing is learned from games other than violence. Finally, myths in educational policy include; class size does not matter, more money necessarily means better quality instruction, and education never changes (De Bruyckere et al., 2015). Characteristics of the Zombie Concept Now that we have defined and provided examples of zombie concepts in education, we return to the zombie concept metaphor. However, to “flesh out” the metaphor a bit further, we describe defining characteristics of zombie concepts including; the zombie environment, zombie longevity, and the power of the zombie herd.

10    G. M. SINATRA and N. G. JACOBSON

Zombie Environment Zombie concepts require an environment wherein they can thrive. In science fiction, zombies exist in a postapocalyptic world, a wasteland that supports their existence. Currently, with the proliferation of fake news, weaponized misinformation, and widespread available of quackery, we view zombie concepts as existing in both the physical and digital spaces. For a brick and mortar example, visitors to the Creationist Museum in Kentucky will be treated to exhibits of humans interacting with dinosaurs, a powerful visual depiction that perpetuates the myth of their contemporary coexistence. With all its strengths and affordances acknowledged, the Internet provides its own zombie friendly zone: an environment rife with chaotically organized and unreliable information. If you query “Did humans and dinosaurs coexist?” you are presented with a lengthy list of sources “confirming” such a link despite the overwhelming scientific evidence to the contrary. Next, we explain how not all zombies are created equal. Longevity Some educational myths are fads which quickly fade away, while others persist for decades. To tease apart the idea of differential staying power, we need to invoke another popularized zombie, that of The Night King in HBO’s Game of Thrones. To provide brief context, the Night King is a fictional zombie character who was the first zombie in this world, and is stronger and more robust than other zombies. Zombie staying power, or longevity, is of vital importance for it highlights the intractability of particular zombie concepts which underlies their role in education. An example of a myth that has persisted in education and popular culture is the notion that while some people are naturally gifted in mathematics, others make the claim that they struggle because they are just not “a math person.” The myth of dichotomous categories (there are math people and non-math people) is a persistent and destructive myth. According to Miles Kimball and Noah Smith (2013), long time mathematics educators, “For high-school math, inborn talent is much less important than hard work, preparation, and self-confidence” (p. 10). Recent work by Dweck and others on growth mindsets suggests that a belief that learning is incremental and that intelligence is not fixed can make a difference in academic outcomes, including in subjects such as mathematics (Rattan, Savani, Chugh, & Dweck, 2015). For students who shift their “math identity” and adopt a growth mindset, improvement in academic outcomes is possible no matter what the starting point, according to Allen and Schnell (2016). Perhaps the rise in appreciation of a growth mindset will finally destroy the longevity of the “math person” myth. However, it is very difficult to extinguish zombies of such strength.

Zombie Concepts in Education    11

Herds Educational myths can gain strength from and are reified by their relation to other zombie concepts. Often zombie concepts group together and travel in “herds” increasing their danger, even allowing for new and emergent zombie concepts. As noted, De Bruyckere et al. (2015) documented that educational myths cluster into herds related to neuromyths, myths about technology in education, and myths about educational policy. Macdonald, Germine, Anderson, Christodoulou, and McGrath (2017) provide empirical support for this herd organization of zombie concepts. Their factor analysis of 22 commonly endorsed educational myths reduced to three herds: (a) myths related to neuroscience, (b) myths related to learning/ education, and (c) myths related to motor coordination. From this evidence, we argue that a zombie herd is exponentially more dangerous than a single zombie because the individual zombies within the herd reinforce and reify one another, potentially resulting in emergent zombie concepts. For example, in MacDonald et al. (2017) endorsement of learning styles, right-brain/left-brain learners, use of 10% of the brain, and the Mozart effect as true all loaded on the “classic: neuromyths factor” (p. 5). This indicates myths “herd” together suggesting that an attempt to correct any one of these incorrect beliefs would be challenging due to its interconnection with other myths. Dole and Sinatra (1998) argued that misconceptions that are strongly connected to other ideas are much harder to overcome. Summary We define zombie concepts as persistent myths that enjoy widespread popular support despite little empirical support and undergird thinking about education. Zombie concepts thrive in the right environment, can enjoy long term staying power, and benefit from herding with other zombie concepts, making them more resistant to eradication. Next, we turn to education and societal and institutional forces that perpetuate these myths in both formal and informal learning environments. WHY ARE ZOMBIE CONCEPTS SO APPEALING? Zombie concepts persist in part because of their widespread appeal. We posit that there are three major factors that contribute to their appeal; (a) zombie concepts are informed by perceptions of reality, not reality; (b) zombie concepts offer compelling, intuitively appealing, comprehensible explanations; and (c) zombie concepts proffer simple solutions to intractable

12    G. M. SINATRA and N. G. JACOBSON

problems. Sadly, these solutions, like snake oil, are quackery, but we will return to this point in a later section. Zombie Concepts Are Informed by Perceptions of Reality, Not Reality The Earth appears flat to someone walking on its surface. Humans tend to explain their reality via perceptions of reality which may or may not align with evidence (Shtulman, 2017). The developmental literature explains that incorrect understandings are natural and privilege perception, while scientific explanations require instruction and a change in individuals’ thinking (Shtulman, 2017). Educational zombie concepts are intuitive ideas that resonate with our experiences of the world. Educators who endorse learning styles may adhere to this belief because they correctly observe students in their classroom have different preferences for interacting with information. It seems to make sense that some students learn visually, and some students learn auditorily, and still others learn kinesthetically because students do have preferences for how to interact with information. Some learners prefer to sit and listen to a lecture while others prefer to read the information on their own. Teachers know that some students are more physically active than others and they prefer to keep moving rather than sit still for long periods. These preferences are indeed real, which is why the idea of learning styles is so resonant, teachers have repeatedly observed these preferences in their students. Zombie Concepts Offer Compelling, Intuitively Appealing, Comprehensible Explanations Because students express preferences for interacting with information, it seems to follow logically, that students learn differently depending on their style. Unfortunately, the nature of learning is much more complex than this simple explanation suggests. The disconnect here comes from a misunderstanding of what it means to prefer a modality versus to learn differently through a modality. A learning preference is simply a preferred modality for experiencing instruction. These preferences are as real as preferences for different literary, music, or film genres. Some people like romance novels, others prefer science fiction. Some people listen to rap music, others enjoy classical. However, and this is the key, the romance reader does not process printed text and comprehend meaning differently than the sci-fi reader. They both process the letters on the page using the same visual and

Zombie Concepts in Education    13

cognitive mechanisms. They both construct meaning by drawing on their background knowledge, and they both remember the plot using similar memory mechanisms. The rap aficionado and our classical music fan process sounds through their auditory systems similarly, evoke emotions from the music using the same parts of the brain, and later whistle their new favorite tune using the same mechanisms of recall. So, too, do learners with typical senses and no disabilities process instructional materials through the same modality using the same learning processes. There is no evidence “visual learners” have different visual processing systems than “auditory learners.”1 It is important to note that there is much more to learning than processing information. The presentation of instructional information may affect students’ willingness to engage and motivation to learn. One student may not have the motivation to listen to an hour-long lecture while another may, which will definitely result in different learning outcomes. Preferences among learners for how to receive instructional information are important and educators should acknowledge these as part of the learning experience and should take students’ preferences for engaging with content into account. However, the learning that occurs when two students are equally motivated, and have similar background knowledge and opportunity to learn does not differ by “style” of learning in any sense of that word. Zombie Concepts Offer Solutions to Intractable Problems When perceptions are used to make decisions rather than evidence, a disconnect is bound to occur. Zombie concepts provide easy to adopt “solutions” to intractable issues in education. An intractable education problem is highlighted by the question, “Why do some students succeed in school while others struggle?” The answer to this question is exceedingly complex and multilayered. Not all students have equal educational opportunities in school or out of school, many are homeless, many more live in poverty (i.e., receive free or reduced lunch), live in stressful home environments; some have diagnosed or undiagnosed health issues, learning disabilities, or motivational challenges. Struggling students may have experienced poor or inadequate instruction, or they may lack support from adults, they may speak a first language other than that of their teacher and receive little to no instruction in their native language. Some are being bullied or have unidentified emotional or mental health challenges. Some students may be struggling due to a mix of personal and contextual factors where some are known and some are unknown to their teachers. The causes of students’

14    G. M. SINATRA and N. G. JACOBSON

struggle to succeed in school are not often easy to identify and even once identified, they are often not easily solved. A quintessential zombie concept such as learning styles offers an easy, comprehensible, relatable, simple solution in the form of: students can be categorized into types of learners with a simple survey administered in a few minutes, instruction can be matched to the learning style of the student, and the challenge of differential success can be addressed by modifying instruction to match learners’ styles. This idea is clearly appealing and if it were true, our jobs as educators would be much easier. The problem is that even when students are provided instruction that matches their preferred style, they do not show greater achievement (Pashler, McDaniel, Rohrer, & Bjork, 2008). Summary Zombie concepts resonate with our experience and that is a large part of their appeal. As such, they offer compelling, intuitive ideas which are easy to understand and provide explanations that promise simple solutions to intractable and enduring educational problems. Poverty, for example, is a major predictor of challenges for student success and obviously impenetrable to quick and easy solutions (Engle & Black, 2008). Racial inequity is another clear contributor to the achievement gap, and again, this is much harder to address than labeling students as visual or auditory learners. Students with learning disabilities such as ADHD and dyslexia, can be very successful but only if they are properly diagnosed and if they receive the instructional adaptations they need to flourish (Blachman, 2013). Again, it is easier to have students fill out a learning styles worksheet than to comprehensively diagnose and remediate specific learning disabilities. Sadly, zombie concepts excuse us from addressing persistent educational issues with costly, time intensive, but more effective solutions such as providing one-on-one classroom aides to students with learning disabilities or providing equal access and equal opportunity for all students to quality instruction that meets their needs. HOW DO ZOMBIE CONCEPTS ENDURE? It is understandable why myths are appealing, but how do they endure? Unfortunately, it is not just the general public that adheres to myths. Several studies have shown that educators endorse myths at the same rate as members of the public (Dekker, Lee, Howard-Jones, & Jolles, 2012; Losh & Nzekwe, 2011; Macdonald et al., 2017; Pashler et al., 2008). Education myths are no exception. The learning styles myth enjoys the greatest levels

Zombie Concepts in Education    15

of acceptance with 76% of educators endorsing this zombie concept according to one study (Simmonds, 2014). What accounts for the persistence and endurance of these ideas among members of the general public and educators? Zombie concepts endure because they have psychological and emotional resonance and they connect with views of ourselves. They also endure because individuals are motivated to have them persist. Zombie Concepts Have Psychological Resonance In addition to resonating with our observations of learners, there are other reasons why these appealing but unsupported concepts are hard to counter. Humans tend to apply heuristic judgments when evaluating data and evidence. Heuristic judgments are “snap” judgments that employ System 1 processing which is more intuitive and less reflective (Kahneman, 2011). If educators use heuristics to judge whether some students seem to learn differently than others, it would be easy to recall a fidgety student who was a struggling learner and the availability of this confirming example makes the idea that such students learn best kinesthetically seem reasonable. Most teachers have likely seen young students reverse letters as it is very common. If a student is a struggling reader, it might be easy to recall an instance where that student reversed letters, providing a confirming, yet incorrect notion of dyslexia. An initial heuristic judgment, when combined with the confirmation bias, or the bias toward the privileging of confirmatory evidence (Nickerson, 1998), reinforces the initial snap judgment. In addition to heuristic judgments, individuals also have dispositions which impact their reasoning. Stanovich (1999) defines dispositions as ‘‘relatively stable psychological mechanisms and strategies that tend to generate characteristic behavioral tendencies and tactics’’ (p. 157). In contrast to heuristic judgments, dispositions can function on the System 2 or reflective level which evoke intentional approaches to thinking and problem solving. Research has shown that individual differences in dispositions relate to problem solving and reasoning. For example, studies have shown that individuals who are disposed towards more open-minded reasoning strategies, or show greater willingness to engage with new information, or enjoy effortful thinking and problem solving (Cacioppo & Petty, 1982), or have greater comfort with ambiguity (Webster & Kruglanski, 1994) show advantages in problem-solving performance and acceptance of scientific explanations (Sá, West, & Stanovich, 1999; Sinatra, Southerland, McConaughy, & Demastes, 2003). These individual differences in dispositions towards thinking and reasoning can play a role in how one reasons about zombie concepts. Individuals may not be motivated to put the time and effort into solving the complex riddle of a students’ apparent underperformance or

16    G. M. SINATRA and N. G. JACOBSON

keep an open mind to alternative explanations for a students’ apparent difficulties, thus allowing the zombie concept to prevail. A growing body of research suggests that individual differences in epistemic cognition (Greene, Sandoval, & Bråten, 2016) reflect differences in willingness and interest in engaging in critical thinking; the type of thinking necessary to attack complex problems in education. The emerging field of epistemic cognition (EC) describes “how people acquire, understand, justify, change, and use knowledge in formal and informal contexts” (Greene et al., 2016, p. 1). Even devoted educators who care deeply about their students, are often overworked individuals who lead busy and overcommitted lives. Thus, whether individuals are disposed to think critically or not, educators may simply not have the time to devote to consuming the most up-to-date research or investigating the nature of a learning problem. In short, zombie concepts fill the void left by unwillingness, disinterest, or simply the pressures of limited time and energy to address the complexity posed by many educational problems. Zombie Concepts Have Emotional Resonance Individuals who endorse learning styles often do so with great passion and conviction. In our own teaching, we have noticed that students in preservice and in-service education programs are reluctant to let go of the idea that learning styles lack empirical support. The idea may be held dearly due to the personal experiences observing struggling students, or they have learned about learning styles in K–12 or higher education courses from respected teachers. One educator calling into a webinar explaining how learning styles lacked empirical support noted that he really did not care what the research shows, because he knew that he had observed students with different learning styles in his classroom. Likely, this teacher was correctly observing differences in his students’ learning preferences, which could explain his certainty and passion. Teachers such as this one should be applauded for their keen powers of observation and their commitment to meeting their students’ learning needs. However, teachers like this one, who are motivated to help their students, would be better served by evidence-based models of instruction rather than myths and educational quackery. Zombie Concepts Resonate With Our Views of Ourselves as Learners Often simple solutions make intuitive sense and offer “explanations” not only about students, but about ourselves (e.g., “Oh, that’s why I’m not so

Zombie Concepts in Education    17

great at math, I’m an auditory learner and I wasn’t taught in my preferred style.”). Zombie concepts may provide a convenient explanation for individuals who experienced difficulties in school or learning themselves. Many individuals struggle through K–12 education for a variety of reasons as noted. The difficulties certainly are real but the cause was unlikely to have been a mismatch between instructional modality and preferred learning style. It is not unusual to hear proponents of learning styles note in defending their position, “I’m a visual learner, and I just know I would have done better in math if my teacher had adapted instruction to meet my style.” Such a claim is partly correct in that any student is likely to have better learning outcomes if a teacher had more effectively met their particular learning needs. Again, students are best served by using effective instructional strategies and approaches that actually support learners’ needs. WHY ZOMBIE CONCEPTS WON’T DIE We have argued that zombie concepts endure by exploring how they resonate with individuals psychologically, emotionally, and personally. However, institutional and social-cultural forces also serve to perpetuate zombie concepts in formal, nonformal, and informal learning environments (Eraut, 2000; Schwier, 2012). Formal learning environments are those with a deliberately organized curriculum led by an instructor who may award credit based on external standards (Eraut, 2000). Formal contexts definitely serve to perpetuate myths. However, much learning about zombie concepts occurs outside of these formal contexts. Nonformal and informal learning environments such as museums, zoos, or the Internet are those where learning occurs either implicitly, reactively, or deliberately (Eraut, 2000; Schwier, 2012). Much of the learning about zombie concepts takes place in informal settings, such as Internet searches. Implicit learning may occur when individuals are learning about a vaccine’s risks while seeking information about school policies without their conscious awareness. This same topic could also be studied in a reactive fashion, for example, by reading an article about vaccines in the New York Times along with other articles on economics, public policy, and so on. Consulting the Internet for an answer to a specific question such as, “Do vaccines cause autism?” would be defined as deliberative learning because the individual planned the search, intentionally engaged with the problem, and actively searched for an answer. Next, we examine forces and actors at play in both formal and informal spaces that support the perpetuation of zombie concepts.

18    G. M. SINATRA and N. G. JACOBSON

Formal Learning Spaces Teacher Education Educational zombie concepts are ubiquitous, and influence the thinking of educators, students, and the public the world over (Dekker et al., 2012; Furnham & Hughes, 2014; Losh & Nzekwe, 2011; Macdonald et al., 2017). For instance, Dekker et al. (2012) documented myth endorsement in UK and Dutch teachers finding that even among teachers interested in teacher education and neuroscience, a full 49% of myths were endorsed as true. Macdonald et al. (2017) highlighted the differences between endorsement of myths among the public, educators, and those experienced in neuroscience, finding that myth endorsement among these groups was surprisingly homogeneous. Additionally, Furnham and Hughes (2014) found that even among those who study psychology, myth endorsement is alarmingly high, with only a small difference existing between psychology students and the public. Unfortunately, there is evidence to suggest that zombie concepts are perpetuated in formal learning contexts. Recall that Losh and Nzekwe (2011) found that preservice teachers’ beliefs about pseudoscience were comparable with those of the comparably educated American adults, and individuals anecdotally report learning about education myths in high school. Rather than debunking education myths, formal schooling often serves to support their perpetuation. Publications Recently, Philip Newton (2015) analyzed the literature in ERIC and PubMed finding that an overwhelming majority (89%) of studies directed at explaining student learning in higher education implicitly or directly endorsed the use of learning styles (Newton, 2015). It is not surprising to find myth endorsement in education publications since many educators themselves believe educational myths. These educational myths are so ingrained into education culture that educators teach them to students, educators teach them to teachers, and educators write about them in education journals. Informal Learning Environments Popular Culture Popular culture plays a role in both bringing attention to scientific findings but also to distorting these same findings (Lakoff & Johnson, 1980). For example, Muela and Abril (2014) analyzed student conceptions of genetics and compared it with the depiction of genetics in film. The authors concluded that cinema can create, support, or reify misconceptions about

Zombie Concepts in Education    19

genetics that later become learning obstacles for students in science class. In their article, the authors highlight films such as The 6th Day, Jurassic Park, Spiderman, X-Men, The Island, and Fantastic Four as films that depict genetic misconceptions. Another educational myth, that we only use 10% of our brain, has been depicted in both film and television—especially via the 2011 film Limitless. Here we have an important confluence of popular culture including celebrity and film reinforcing a myth. Bradley Cooper, a popular actor, was the lead in Limitless. We know from the persuasion literature that source effects such as celebrity endorsement increases adoption of positions (Erdogan, 1999; Petty & Cacioppo, 1986). Watching a popular culture film arguably results in implicit learning, if the viewer is watching for entertainment or fun instead of approaching the film with specific questions in mind about the content. Commercialization Large sums of money can and will continue to be made by taking advantage of parents’ concern for their child’s education. Petroff (2017) reported that U.S. parents on average will spend $58,424 dollars on their child’s education between grade school and the end of an undergraduate degree with public school parents spending $39,544 and private school parents spending $74,112. This number includes tuition and transportation in addition to technology and other costs. Relevant to our discussion here are the discretionary funds spent by parents on educational toys, games, and other products. In a 2015 study of 1000 parents, they planned to spend between $873 and $1,124 dollars purchasing educational technology for their children (Rubicon Project, 2015). If we contextualize this value by multiplying it by the approximate number of students (IES, 2017) in the U.S. education system (50 million in public elementary and secondary schools, 5 million in private elementary and secondary schools, 20 million in colleges and universities) we see that this is at least a $7.5 billion dollar industry. This context highlights the role of commercial interests in perpetuating zombie concepts. One classic example is the zombie concept that listening to classical music increases IQ in babies, commonly referred to as the Mozart Effect (Rauscher et al., 1993). The Walt Disney Company purchased the Baby Einstein franchise in 2001 for $25 million (Mook, 2002). Yet again we see the power of popular culture in that Baby Einstein was featured on the Oprah Winfrey Show, The Today Show, USA Today, and the 2007 State of the Union Address. One other example related to both popular culture and the power of commercial interest is the rise of brain training platforms like Lumosity. Lumosity is an online platform that claimed to improve cognitive performance, brain performance, and possibly fight off the effects of Alzheimer’s

20    G. M. SINATRA and N. G. JACOBSON

through the use of online games (Hiltzik, 2016). Recently, Bainbridge and Mayer (2018) conducted a randomized experiment to assess the claims that Lumosity’s flexibility and attention games increase cognitive skills compared to an inactive group of learners. Their results indicated that the claims made by Lumosity were not empirically supported, a conclusion echoed by the Federal Trade Commission who fined Lumosity $2 million dollars for deceptive advertising due to “a lack of scientific evidence” (Dickey, 2016). As a testament to the staying power of zombie concepts and the force of commercial interest, consider the fact that a company whose product does not do what it claims and is fined, is still in business. Summary Zombie concepts, educational myths, and quackery perpetuate through formal, nonformal, and informal learning environments. In formal learning spaces, experts, educators, and students endorse incorrect myths about education and these myths are reinforced from elementary school, to teacher education programs, to the highest levels of education. In informal learning spaces, the Internet, popular culture, and commercialization each play a role in the perpetuation of educational myths. This analysis, coupled with our discussion of individual level factors that influence myth endorsement, sheds light on how and why zombie concepts stay perpetually alive in education. STAVING OFF THE ZOMBIE CONCEPT APOCALYPSE We have argued that zombie concepts are enduring, unsupported ideas that cannot be easily extinguished. That sounds discouraging. However, we do not mean to imply that there is no defense against the hungry zombie hoard. There are many active measures that can be taken to stave off the zombie concept apocalypse in education. We suggest several specific steps that should be taken and elaborate on these below. Fact Checking for Educational Quackery First, there are a variety of resources now available to identify fake news and to fact check information including FactCheck.org and allsides.com (Pennycook & Rand, 2018). However, we know of no sites where parents and educators can quickly and easily fact check education news or fact check claims. We do know that if you Google “Do we only use 10% of our brains?” plenty of

Zombie Concepts in Education    21

“evidence” will emerge supporting this myth. The What Works Clearinghouse (2017; https://ies.ed.gov/ncee/wwc/) maintains an excellent database of research-tested instructional and curricula approaches and we encourage its use as a resource. The National Education Policy Center (2018) also has a webpage (https://nepc.colorado.edu/) which can serve as resource. These are great sources if one knows how to locate them. However, we know of no easy way for parents and teachers to fact check specific educational claims. We encourage the educational research community to create such resources. Deconstruction of False Arguments and Claims There are also excellent resources currently available for debunking false arguments and claims about scientific myths and controversies around issues such as vaccine safety, genetically modified organisms (GMOs), and humans’ role in climate change (Cook, van der Linden, Maibach, & Lewandowsky, 2018). Cook and colleagues (2018) provide an argumentation primer on how to push back against false information. Their guide illustrates argumentation moves and stratagems to counter ad hominin attacks. Lewandowsky’s research team produced the “Debunking Handbook” which provides a step by step guide to undermining false claims against climate change (Cook & Lewandowsky, 2011). These resources can be readily adopted for pushing back against misinformation, fake news, and quackery in education. Refutational Approaches Refutational texts feature a three-part structure designed to: (a) elicit a misconception, (b) provide a cue that the idea is incorrect, and (c) provide the correct conception along with explanatory evidence. This text format has been shown to impact the correction of misconceptions in science (Guzzetti, 1993; Sinatra & Broughton, 2011; Tippett, 2010). Recently, Aguilar, Polikoff, and Sinatra (2019) demonstrated that the refutation text approach can be used successfully to overcome the misconceptions held by the general public about educational policy. A short refutation text was designed to target misconceptions about Common Core (that it was an Obama era policy and Common Core is prescriptive). Results indicated that support for Common Core could be increased when misconceptions about the policy were reduced through reading the refutation text. Kowalski and Taylor (2017) have demonstrated that refutational style of teaching (deliberately upending misconceptions about psychology through direct refutational content in introductory psychology classes), does work to lower

22    G. M. SINATRA and N. G. JACOBSON

misconceptions about psychology concepts. Again, these strategies and approaches could be adopted to refute misconceptions about misinformation in education. Beyond Myth Busting Some research has shown that training in the content regarding the myth alone does not necessarily counter the myth (Macdonald et al., 2017). Not only is direct instruction in the content not guaranteed to upend the myth, backfire effects can result (Nyhan & Reifler, 2018; Trevors, Muis, Pekrun, Sinatra, & Winne, 2016; Van Loon, Dunlosky, Van Gog, Van Merriënboer, & De Bruin, 2015). A backfire effect is the ironic strengthening of a misconception when provided with contradictory information (Cook & Lewandowsky, 2011). This “doubling down” on original conceptions in the face of conflicting data seems to be especially pernicious for controversial topics. To avoid backfire effects when confronting educational myths and misinformation, we believe it is important to augment the refutational information with a clear and compelling explanation of the correct conception (Kendeou, Braasch, & Bråten, 2016). We, like many others (Chinn, Rinehart, & Buckland, 2014; Greene et al., 2016; Sinatra & Hofer; 2016) have called for critical thinking instruction in K–12 and higher education. Critical thinking is a key component of digital literacy instruction which supports students’ understanding of how to check source information, fact check, evaluate evidence, and weigh issues and arguments (Greene, Seung, & Copeland, 2014). Critical thinking has been shown to be necessary to adequately evaluate myths (Bensley & Lilienfeld, 2017). In addition to critical thinking instruction, we believe it is necessary to promote and support plausibility judgments (Lombardi & Sinatra, 2018). In a recent chapter entitled “Don’t Believe Everything You Think,” Lombardi and Sinatra argue that making a judgment about the plausibility of information is a critical step that should be added to critical evaluation of scientific information. Sinatra and Lombardi (2019) recently argued that plausibility judgments must be incorporated into digital literacy/critical thinking instruction. Many claims (such as rising ocean acidity) cannot be easily tested by lay individuals. So too, it is unrealistic for parents and teachers to test the veracity of most educational claims directly. Thus, K–12 and higher education must enhance critical thinking and digital literacy skills with the ability to judge the plausibility of evidence supporting claims about education. In our own teaching with undergraduate college freshman, we employed a technique we believe has promise. We had students conduct their own empirical research on an educational myth of their own choice. By directly

Zombie Concepts in Education    23

engaging in research testing how widespread their college student peers subscribed to education myths and analyzing the data for themselves, they were able to engage in the type of deep metacognitive and self-regulated engagement that we and others have argued promote epistemic conceptual change (Sinatra & Chinn, 2011). CONCLUSIONS We have defined zombie concepts as persistent myths about education that are hard to extinguish. These myths find support in individuals’ psychological, emotional, and personal commitments to these ideas. Aspects of formal and informal learning environments also provide safe haven to these ideas. We still assert that resistance is not futile, but rather there are a number of approaches that are being used to confront the broader fake news phenomena and to shift attitudes and conceptions about scientific ideas that can be readily adapted to confront the pernicious spread of education myths. We urge all educators to be ever vigilant and avoid falling victim to these ideas, as you will only spread them. We also urge educators and education researchers to rise up and intellectually combat the education zombie concept apocalypse. NOTE 1. There are real differences in how students with visual impairment or dyslexia should experience instruction to optimize their learning potential. However, labeling a student with a visual impairment or a student with dyslexia “an auditory learner” does a grave disservice to that student who instead should be offered research-based effective learning solutions to address their specific learning needs.

REFERENCES Aguilar, S., Polikoff, M., & Sinatra, G. M. (2019). Refutation texts: A new approach to changing public misconceptions about education policy. Educational Researcher, 48(5), 263–272. doi:10.3102/0013189X19849416 Allen, K., & Schnell, K. (2016). Developing mathematics identity. Mathematics Teaching in the Middle School, 21(7), 398–405. Bainbridge, K., & Mayer, R. E. (2018). Shining the light of research on Lumosity. Journal of Cognitive Enhancement, 2(1), 43–62.

24    G. M. SINATRA and N. G. JACOBSON Bensley, D. A., & Lilienfeld, S. O. (2017). Psychological misconceptions: Recent scientific advances and unresolved issues. Current Directions in Psychological Science, 26(4), 377–382. Blachman, B. A. (2013). Foundations of reading acquisition and dyslexia: Implications for early intervention: Abingdon, England: Routledge. Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42(1), 116–131. Chinn, C. A., Rinehart, R. W., & Buckland, L. A. (2014). Epistemic cognition and evaluating information: Applying the AIR model of epistemic cognition. In C. Chinn, R. Rinehart, & L. Buckland (Eds.), Processing inaccurate information: Theoretical and applied perspectives from cognitive science and the educational sciences (pp. 425–453). Cambridge, MA: The MIT Press. Cook, J., & Lewandowsky, S. (2011). The debunking handbook. St. Lucia, Australia: University of Queensland. Cook, J., van der Linden, S., Maibach, E., & Lewandowsky, S. (2018). The consensus handbook: Why the scientific consensus on climate change is important. https://www. doi.org/10.13021/G8MM6P De Bruyckere, P., Kirschner, P. A., & Hulshof, C. D. (2015). Urban myths about learning and education. Walthman, MA: Academic Press. Dekker, S., Lee, N. C., Howard-Jones, P., & Jolles, J. (2012). Neuromyths in education: Prevalence and predictors of misconceptions among teachers. Frontiers in Psychology, 3, 429. Dickey, M. R. (2016). Lumosity “Brain Training” app maker to pay $2 million settlement to FTC. Retrieved from https://techcrunch.com/2016/01/06/ lumosity-brain-training-app-maker-to-pay-2-million-settlement-to-ftc/ Dole, J. A., & Sinatra, G. M. (1998). Reconceptualizing change in the cognitive construction of knowledge. Educational Psychologist, 33(2–3), 109–128. Engle, P. L., & Black, M. M. (2008). The effect of poverty on child development and educational outcomes. Annals of the New York Academy of Sciences, 1136(1), 243–256. Eraut, M. (2000). Non-formal learning and tacit knowledge in professional work. British Journal of Educational Psychology, 70(1), 113–136. Erdogan, B. Z. (1999). Celebrity endorsement: A literature review. Journal of Marketing Management, 15(4), 291–314. Furnham, A., & Hughes, D. J. (2014). Myths and misconceptions in popular psychology: Comparing psychology students and the general public. Teaching of Psychology, 41(3), 256–261. Greene, J. A., Sandoval, W. A., & Bråten, I. (2016). Handbook of epistemic cognition. Abingdon, England: Routledge. Greene, J. A., Seung, B. Y., & Copeland, D. Z. (2014). Measuring critical components of digital literacy and their relationships with learning. Computers & Education, 76, 55–69. Guzzetti, B. J. (1993). Promoting conceptual change in science: A comparative meta-analysis of instructional interventions from reading education and science education. Reading Research Quarterly, 28(2), 116–159.

Zombie Concepts in Education    25 Hiltzik, M. (2016). If you weren’t smart enough to know Lumosity was making bogus claims, the FTC has your back. Retrieved from http://www.latimes.com/business/ hiltzik/la-fi-mh-if-you-weren-t-smart-enough-20160106-column.html IES. (2017). Back to school statistics. Retrieved from https://nces.ed.gov/fastfacts/ display.asp?id=372 Kahneman, D. (2011). Thinking, fast and slow. London, England: Macmillan. Kendeou, P., Braasch, J. L., & Bråten, I. (2016). Optimizing conditions for learning: Situating refutations in epistemic cognition. Journal of Experimental Education, 84(2), 245–263. Kimball, M., & Smith, N. (2013). The myth of ‘I’m bad at math’. The Atlantic. Retrieved from https://www.theatlantic.com/education/archive/2013/10/the -myth-of-im-bad-at-math/280914/ Kirschner, P. A., & Neelen, M. (2018, April 3). Why myths are like zombies. Retrieved from https://3starlearningexperiences.wordpress.com/2018/04/03/why-myths -are-like-zombies/ Kowalski, P., & Taylor, A. K. (2017). Reducing students’ misconceptions with refutational teaching: For long-term retention, comprehension matters. Scholarship of Teaching and Learning in Psychology, 3(2), 90–100. Lakoff, G., & Johnson, M. (1980). The metaphorical structure of the human conceptual system. Cognitive Science, 4(2), 195–208. Lévi-Strauss, C. (1995). Myth and meaning: Cracking the code of culture. New York, NY: Schocken Books. Lombardi, D. & Sinatra, G. M. (2018). Don’t believe everything you think: Reappraising judgments about conceptions. In T. Amin & O. Levrini (Eds.), Converging perspectives on conceptual change: Mapping an emerging paradigm in the learning sciences (pp. 237–244). New York, NY: Routledge. Losh, S. C., & Nzekwe, B. (2011). Creatures in the classroom: Preservice teacher beliefs about fantastic beasts, magic, extraterrestrials, evolution and creationism. Science & Education, 20(5–6), 473–489. Macdonald, K., Germine, L., Anderson, A., Christodoulou, J., & McGrath, L. M. (2017). Dispelling the myth: Training in education or neuroscience decreases but does not eliminate beliefs in neuromyths. Frontiers in Psychology, 8, 1314. https://doi.org/10.3389/fpsyg.2017.01314 Mook, B. (2002). In a ‘tot’-anic size ’01 deal, Disney buys Baby Einstein. Retrieved from https://www.bizjournals.com/denver/stories/2002/03/04/focus9.html Muela, F. J., & Abril, A. M. (2014). Genetics and cinema: Personal misconceptions that constitute obstacles to learning. International Journal of Science Education, Part B: Communication and Public Engagement, 4(3), 260–280. National Education Policy Center, University of Colorado Boulder. (2018). Funding special education: Charting a path that confronts complexity and crafts coherence. Retrieved from https://nepc.colorado.edu Newton, P. M. (2015). The learning styles myth is thriving in higher education. Frontiers in Psychology, 6, 1908. https://doi.org/10.3389/fpsyg.2015.01908 Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2(2), 175–220.

26    G. M. SINATRA and N. G. JACOBSON Nyhan, B., & Reifler, J. (2018). The roles of information deficits and identity threat in the prevalence of misperceptions. Journal of Elections, Public Opinion and Parties, 29(2), 1–23. Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008). Learning styles: Concepts and evidence. Psychological Science in the Public Interest, 9(3), 105–119. Pennycook, G., & Rand, D. G. (2018). Crowdsourcing judgments of news source quality. Retrieved from https://www.researchgate.net/profile/Gordon_Pennycook/ publication/323460988_Crowdsourcing_Judgments_of_News_Source_Quality/ links/5c06a3e292851c6ca1fd5c82/Crowdsourcing-Judgments-of-News -Source-Quality.pdf Petroff, A. (2017). How much do parents spend on education? Retrieved from http:// money.cnn.com/2017/06/29/pf/education-costs-hong-kong-tuition/index .html Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In Communication and persuasion (pp. 1–24). New York, NY: Springer. Rattan, A., Savani, K., Chugh, D., & Dweck, C. S. (2015). Leveraging mindsets to promote academic achievement: Policy recommendations. Perspectives on Psychological Science, 10(6), 721–726. Rauscher, F. H., Shaw, G. L., & Ky, C. N. (1993). Music and spatial task performance. Nature, 365(6447), 611. https://www.doi.org/10.1038/365611a0 Rubicon Project. (2015). Back-to-school consumer pulse poll. Los Angeles, CA: Author. Retrieved from http://rubiconproject.com/wp-content/uploads/2015/07/ Back-to-School-Consumer-Pulse-Full-Results-for-posting-FINAL.pdf Sá, W. C., West, R. F., & Stanovich, K. E. (1999). The domain specificity and generality of belief bias: Searching for a generalizable critical thinking skill. Journal of Educational Psychology, 91(3), 497–510. Schwier, R. A. (2012). Comparing formal, non-formal, and informal online learning environments. In The next generation of distance education (pp. 139–156). New York, NY: Springer. Shtulman, A. (2017). Scienceblind: Why our intuitive theories about the world are so often wrong: London, England: Hachette. Simmonds, A. (2014). How neuroscience is affecting education: Report of teacher and parent surveys. London, England: Wellcome Trust. Sinatra, G. M. (2018, April). Evaluating sources of scientific evidence in the post truth era. Paper presented at the American Educational Research Association, New York, NY. Sinatra, G. M., & Broughton, S. H. (2011). Bridging reading comprehension and conceptual change in science education: The promise of refutation text. Reading Research Quarterly, 46(4), 374–393. Sinatra, G. M., & Chinn, C. A. (2011). Thinking and reasoning in science: Promoting epistemic conceptual change. Critical Theories and Models of Learning and Development Relevant to Learning and Teaching, 1, 257–282. Sinatra, G. M., & Hofer, B. K. (2016). Public understanding of science: Policy and educational implications. Policy Insights from the Behavioral and Brain Sciences, 3(2), 245–253.

Zombie Concepts in Education    27 Sinatra, G. M., Kienhues, D., & Hofer, B. K. (2014). Addressing challenges to public understanding of science: Epistemic cognition, motivated reasoning, and conceptual change. Educational Psychologist, 49(2), 123–138. Sinatra, G. M., & Lombardi, D. (2019). Evaluating sources of scientific evidence and claims in the post-truth era may require plausibility judgments. Manuscript submitted for publication. Sinatra, G. M., Southerland, S. A., McConaughy, F., & Demastes, J. W. (2003). Intentions and beliefs in students’ understanding and acceptance of biological evolution. Journal of Research in Science Teaching, 40(5), 510–528. Stanovich, K. E. (1999). Who is rational?: Studies of individual differences in reasoning. New York, NY: Psychology Press. Tippett, C. D. (2010). Refutation text in science education: A review of two decades of research. International Journal of Science and Mathematics Education, 8(6), 951–970. https://doi.org/10.1007/s10763-010-9203-x Trevors, G. J., Muis, K. R., Pekrun, R., Sinatra, G. M., & Winne, P. H. (2016). Identity and epistemic emotions during knowledge revision: A potential account for the backfire effect. Discourse Processes, 53(5–6), 339–370. https://doi.org/10 .1080/0163853x.2015.1136507 Van Loon, M. H., Dunlosky, J., Van Gog, T., Van Merriënboer, J. J., & De Bruin, A. B. (2015). Refutations in science texts lead to hypercorrection of misconceptions held with high confidence. Contemporary Educational Psychology, 42, 39–48. Webster, D. M., & Kruglanski, A. W. (1994). Individual differences in need for cognitive closure. Journal of Personality and Social Psychology, 67(6), 1049–1062. What Works Clearinghouse,  Institute of Education Sciences, U.S. Department of Education. (2017, October). What Works Clearinghouse: Procedures Handbook (Version 4.0). Retrieved from http://whatworks.ed.gov

This page intentionally left blank.

CHAPTER 3

UNDERSTANDING SUSCEPTIBILITY TO EDUCATIONAL INACCURACIES Examining the Likelihood of Adoption Model Alexandra List The Pennsylvania State University Lisa DaVia Rubenstein Ball State University

ABSTRACT The purpose of this chapter is to consider the nature of inaccurate information in education and to identify the factors that may influence individuals’ adoption of accurate, inaccurate, or misleading information from text. We begin by defining the origins, features, and types of educational inaccuracies that students may encounter. We then introduce the likelihood of adoption

Misinformation and Fake News in Education, pages 29–54 Copyright © 2019 by Information Age Publishing All rights of reproduction in any form reserved.

29

30    A. LIST and L. D. RUBENSTEIN model (LAM) to taxonomize the learner and message related factors that may impact students’ adoption of accurate and inaccurate information, alike. In an initial test of the LAM, hypotheses introduced by this model were largely supported. Students found the provision of more evidence to be more convincing than the provision of less evidence, and students preferred statistical information to anecdotal information. Moreover, these effects manifest regardless of whether students were presented with accurate or inaccurate information, suggesting that both types of information are processed similarly by learners. This work suggests the importance of teaching students to engage in more deliberative processing, rather than using heuristics to evaluate information.

Adopting accurate information regarding educational practices promotes efficient and effective learning, yet students commonly use strategies like highlighting and rereading, which have been found to be either ineffective or less effective than deeper-level strategies, like summarization or elaboration (Dunlosky, Rawson, Marsh, Nathan, & Willingham, 2013). Many teachers consider identifying and matching instructional activities to students’ learning styles to be highly effective (Scott, 2010), despite the fact that Pashler, McDaniel, Rohrer, and Bjork (2008), charged with evaluating the research on learning styles concluded: “Limited education resources would better be devoted to adopting other educational practices that have a strong evidence base” (p. 105). In this chapter, we examine the types of inaccuracies, fallacies, and misconceptions common in education and consider how these may arise. We have three primary goals. First, we integrate literatures from education, persuasion, and psychology to consider how educational inaccuracies are propagated and specifically, which factors may make students more or less susceptible to adopting inaccurate or misleading information from text. Second, we introduce the likelihood of adoption model (LAM) as an initial attempt to taxonomize the learner and message related factors that may make students more or less likely to adopt inaccurate information about education. Finally, we test some of the factors introduced in the LAM in a sample of undergraduate students. PROPAGATING EDUCATIONAL INACCURACIES Within this chapter, we use inaccuracies as a general term to refer to the status of information as inaccurate relative to some normative standards of truth or reality. Literatures in psychology, communications, and rhetorical persuasion provide insights into factors that may make adopting inaccuracies more or less likely for individuals. In the broadest sense, inaccuracies, much like accurate information, are communicated messages disseminated

Understanding Susceptibility to Educational Inaccuracies    31

Figure 3.1  Demonstration of the reciprocal process of disseminating information.

by a source to an audience (see Figure 3.1). Source, message, and audience components interact and inform each other, such that the source considers the intended audience when developing a message and the audience interprets and evaluates not only specific messages but also their sources (Lasswell, 1948). In education, as in other fields, inaccurate messages may originate from a variety of sources and be communicated to a variety of audiences. Education may be thought of as a complex and multidimensional system, serving and beholden to a variety of constituencies, including parents, students, teachers, administrators, and policy makers. Due to its unique societal position, education may be susceptible to inaccuracies targeting or perpetuated by each of these constituencies. For example, learners may have internalized inaccuracies regarding which learning strategy is most effective (Callender & McDaniel, 2009), while taxpayers and policy makers may be susceptible to misinformation regarding how best to leverage available funds to improve schools (Convertino, 2017; Kovacs & Christie, 2010). The variety of stakeholders and constituencies susceptible to educational inaccuracies makes these both voluminous in number and varied in character. These inaccuracies are propagated through a variety of mechanisms, such as popular media outlets, teacher professional development sessions, and informal peer experiences. Given the underlying, influential factors may be different for these different platforms, we choose to focus our examination on how texts may be used to spread inaccurate information. As vehicles for disseminating inaccurate information, texts, in particular are important to consider for at least three reasons. First, unlike conversations, texts constitute reified messages that are deliberately created and potentially revised to communicate a message to some audience, rather than spontaneously produced (Alexander & Fox, 2004). This reified or physical characteristic of texts endows them with a sheen of reliability, potentially making them more likely to be believed. Second, texts are oftentimes

32    A. LIST and L. D. RUBENSTEIN

perceived by learners as authorless entities, divorced from their sources of origin (Graesser, Bowers, Olde, & Pomeroy, 1999; Paxton, 2002; Shanahan, 1992). This stands in contrast to oral communication, which is more likely to provide students with overt or explicit information about message origin. As such, we were interested in the extent to which students attended to author or source information in processing accurate and inaccurate information. Finally, we elected to present students with information via texts because of their predominance as information sources in our society. Indeed, much prior work on reading comprehension and text evaluation has focused on asking students to evaluate the trustworthiness and accuracy of information presented using traditional (i.e., print based) and digital channels, including the Internet (Britt & Aglinskas, 2002; Bråten, Strømsø, & Britt, 2009; List, Alexander, & Stephens, 2017). By examining inaccuracies presented via texts, in this study, we wanted to speak to this prior work examining how students evaluate information in print and online. The subsequent sections, anchored in persuasion literature, explore how characteristics of the source, message, and audience influence an individual’s likelihood of adopting both accurate and inaccurate information. Sources and Inaccurate Information Essential to critically reading and evaluating texts is the fundamental recognition that texts are written by someone for some purpose (Alexander & The Disciplined Reading and Learning Research Laboratory, 2012; Alexander & Fox, 2004). Regarding source characteristics, the literature on persuasion has identified author credibility, affiliation, expertise, cognitive authority, attractiveness, likeability, and perceived objectivity as factors, among others, influencing students’ evaluations and adoption of message content (Berlo, Lemert, & Mertz, 1969; Chaiken, 1980; O’Keefe, 2002; Petty & Cacioppo, 1984; Rieh & Belkin, 1998). Literature from other fields have corroborated the importance of source characteristics in relation to the likelihood that a particular message will be adopted, including author expertise, trustworthiness, relevance, likeability, and similarity to readers (Feng & MacGeorge, 2010; List et al., 2017; Ohanian, 1990; Pornpitakpan, 2004; Wilson & Sherrell, 1993). Individuals may make source-based or author-based evaluations of information, assuming that authors, deemed to be trustworthy, are likely to provide accurate and quality information (Metzger, 2007; Rieh, 2002; Rieh & Belkin, 1998). When rendering source-based evaluations of texts’ trustworthiness, students consider author expertise (i.e., authority and presumed knowledge) and benevolence (i.e., intention to provide accurate

Understanding Susceptibility to Educational Inaccuracies    33

and quality information; Stadtler & Bromme, 2014). In general, inaccuracies may arise as a result of specific source-related factors, either through lack of knowledge or lack of benevolence (i.e., intentional malintent) on the part of the author (Metzger, 2007; Rieh, 2002; Rieh & Belkin, 1998). Delineating a Typology of Source-Related Inaccuracies Presented in Figure 3.2, a lack of author expertise or a lack of positive intent leads to three types of inaccuracies. The first two types arise as a result of authors’ pretense to knowledge without true expertise or quackery. When authors are well-meaning, but lacking in the requisite information, we consider them to be perpetuating pseudoscience; whereas, when authors’ lack of knowledge is further accompanied by malintent or the desire to profit off of providing inaccurate information, we consider them to be perpetuating fraudulence. Indeed, in the framework introduced in Figure 3.2, pseudoscience is distinguished from fraudulence according to author intent, as benevolent or not. When authors hold malintent, while at the same time having the requisite knowledge and expertise to know that the information that they are providing, is, in fact, incorrect, we refer to such inaccuracies as misinformation or “fake news.” The provision of misinformation or “fake news” can be characterized as particularly nefarious as it stems both from authors’ malintent and from their specific knowledge of the contrary. Collectively, these three types of inaccuracies can be contrasted with benevolent and knowledgeable authors providing accurate and quality information intended either to inform or persuade. Moreover, these terms can be contrasted with misconceptions, much examined in the educational psychology literature (Chi, 2005; Kendeou & Van Den Broek, 2005; Sinatra, Brem, & Evans, 2008; Vosniadou, 1994), or cognitive biases (Gilovich, Griffin, & Kahneman, 2002; Pennycook, Ross, Koehler, & Fugelsang, 2017; Tversky & Kahneman, 1974, 1978), explored in cognitive psychology, which arise through students’ own misunderstandings, inaccurate perceptions of, or faulty reasoning about the world.

Figure 3.2  A typology of inaccuracies.

34    A. LIST and L. D. RUBENSTEIN

Sources Within an Education Context Determining the source’s expertise and intent may be challenging within educational contexts. Specifically, the illusion of understanding that is perpetuated in education further makes determinations of expertise in education more difficult to render. Indeed, expertise in education may be more difficult to establish when experts, recognizing the complex and contextual nature of education, may be reluctant to make strong pronouncements regarding what is true (Eisenhart, 2005). To start, as is the case in all domains, there are no doubt experts in education. At the same time, the field of education is rife with teachers considering educational researchers to be disconnected from classroom practice (Boardman, Argüelles, Vaughn, Hughes, & Klingner, 2005; Carnine, 2000) and undergraduates, often incorrectly, considering themselves to be experts on their own learning (Schmidt & Bjork, 1992; Yan, Bjork, & Bjork, 2016). Absent the recognition of an actual authority figure, inaccuracies in education may abound. At the same time, the nature of education may make expertise difficult to determine. In part, this may come from the pendulum swings of educational practice (Alexander, Murphy, & Woods, 1996), wherein teachers witness certain “research-based” practices replaced with other “researchbased” practices every couple of years (e.g., the shift from whole-language to phonics-driven reading instruction; Carnine, 1997, 2000; Stanovich & Stanovich, 1997). Moreover, the inherently individualized nature of education may further serve to perpetuate inaccuracies, by allowing researchbased pronouncements to potentially be refuted by personal or anecdotal experience (Biesta, 2007; Prabhu, 1990; Timperley, 2005). Indeed, Yan and colleagues (2016) demonstrated how strongly students hold the belief that they are unique as learners and that what improves learning for others may be different from what would improve learning for them. Likewise, teachers have been found to endorse colleagues as providing better professional development than college courses or professional journals (Landrum, Cook, Tankersley, & Fitzgerald, 2002). The resulting ambiguity regarding educational expertise may further allow inaccuracies to fester. Message/Text Characteristics Sources communicate their messages using content and methods that may make students’ information adoption more or less likely. Important factors include both the characteristics of the source (i.e., message provider, like author) and message content. As aforementioned, source characteristics are influential; nevertheless, in this chapter we are particularly interested in the role of content-related factors in message absorption, when content is assigned to a seeming credible source.

Understanding Susceptibility to Educational Inaccuracies    35

The most commonly manipulated content related factors pertain to argument quantity and strength (Cacioppo, Petty, & Morris, 1983; McCrudden & Barnes, 2016; Petty & Cacioppo, 1984; Voss & Means, 1991). Petty, Harkins, and Williams (1980) differentiate strong and weak arguments as arguments either providing relevant, statistical information or opinions and quotations. Indeed, the comparison between anecdotal and statistical information has commonly been drawn (Baesler & Burgoon, 1994; Hoeken, 2001). More comprehensively, Hoeken and Hustinx (2003) distinguish between evidence that is anecdotal (i.e., examples or illustrations), statistical (i.e., numerical description of instances), causal (i.e., causal explanation of a relation), or expert (i.e., citing an authority opinion) in nature. However, they found only a distinction between students’ evaluations of anecdotal evidence vis-a-vis statistical, causal, and expert evidence. Other work has further differentiated between students’ evaluations of causal vis-a-vis correlational evidence provided in persuasive texts, with mixed results (Hoeken, 2001; Slusher & Anderson, 1996). Beyond contrasting strong versus weak arguments, students have been found to hold an explanation bias, such that they are more persuaded by information providing an explanation for the relation between two variables (Anderson & Sechler, 1986; Ross, Lepper, Strack, & Steinmetz, 1977). Elsewhere, students have been found to be more persuaded by message comprehensibility, more readily adopting information that is easier to understand (Eagly, 1974; Scharrer, Bromme, Britt, & Stadtler, 2012). The comprehensibility provides an example of the interaction between message and reader characteristics (e.g., reading comprehension skills). Audience/Reader Characteristics Beyond source and message factors, a variety of individual reader characteristics may impact students’ adoption of information. These factors may be distinguished as more durable (i.e., persistent learning characteristics) or more situational (i.e., topic and task specific) in nature. Further, within the education context, most readers have considerable personal experiences that may influence their adoption of information. Durable Factors Durable factors include students’ prior beliefs or attitudes about a variety of topics, which have been identified as among the most impactful factors determining belief maintenance, the ease of adoption of belief-consistent information, and the critical evaluation and dismissal of belief-inconsistent content (Ditto & Lopez, 1992; Edwards & Smith, 1996; Kunda, 1990; Lord, Ross, & Lepper, 1979; Nickerson, 1999; Taber & Lodge, 2006). Additionally,

36    A. LIST and L. D. RUBENSTEIN

cognitive factors, like students’ need for cognition, have been found to impact students’ evaluation and adoption of persuasive arguments (Cacioppo et al., 1983; Haugtvedt & Petty, 1992). Situational Factors Nevertheless, leading theories explaining students’ adoption of persuasive information highlight the role of situation-specific individual difference factors. Two specific models articulate the role of individual differences in students’ adoption of persuasive information. Most prominently, the Elaboration Likelihood Model (ELM) describes the circumstances under which students may be more or less likely to engage in issue-relevant thinking or to critically evaluate persuasive content (Petty & Cacioppo, 1984, 1986). The ELM argues that the mechanism of information delivery (i.e., through peripheral or central avenues) influences to what extent learners critically evaluate the information. Information delivered via the central route may be expected to be carefully evaluated or scrutinized, with students expending a great deal of effortful on topic-relevant thinking. Information delivered via the peripheral route may, however, be expected to be evaluated only heuristically, via superficial cues. For instance, students may default to preferring or adopting messages having more, rather than less, arguments when processing information at a superficial or peripheral level. The type of processing that students engage in (i.e., central or peripheral) may be considered to be the individual difference factor, arising from the specific task context, that impacts whether students adopt information or elect not to do so. In a similar vein, Chaiken (1980) introduced the heuristic-systematic model of persuasion, which suggests students use either heuristic or systematic processing when presented with persuasive content. Individuals predominantly use heuristic or low-level processing when confronted with persuasive information, evaluating content based only on superficial cues (Chaiken, 1980). In contrast, systematic processing occurs when students deliberately evaluate and analyze information, expending significant cognitive effort to do so. As such, Chaiken’s (1980) conception of heuristic versus systematic processing can be thought to align with Petty and Cacioppo’s (1984) conception of peripheral vis-a-vis central routes to persuasion. Crucially, across both the ELM and the heuristic-systematic processing model, students’ engagement in superficial or deep-level processing is considered to be the guiding factor impacting their evaluation and adoption of information. The level of processing that students engage in may be triggered by situational or task factors (e.g., time limits; cuing students to whether persuasive information is relevant or not) or by students’ more durable individual difference characteristics (e.g., relevance, interest in a topic, motivation; Chaiken & Maheswaran, 1994; Petty & Cacioppo, 1990). At the same time, reader factors impacting message adoption have often

Understanding Susceptibility to Educational Inaccuracies    37

been found to be in interaction with source and message characteristics (Cacioppo, Petty, Kao, & Rodriguez, 1986; Chaiken, 1980; McCrudden & Barnes, 2016; Petty, Cacioppi, & Kasmer, 2015; Slater & Rouner, 1996), when students engage in both superficial/heuristic and deep-level/systematic processing. This literature suggests that the consideration of source or author factors occurs primarily when students engage in superficial or heuristical-level processing (Berlo et al., 1969; Chaiken, 1980; O’Keefe, 2002; Petty & Cacioppo, 1984; Rieh & Belkin, 1998). When students more systematically evaluate information, they may, rather, evaluate actual source content (Chaiken, 1980; Petty & Cacioppo, 1984, 1986). When students do not have the requisite knowledge or resources to critique texts (Kienhues, Stadtler, & Bromme, 2011), they may rely more heavily on source-based or author-based evaluations of information, assuming that authors, deemed to be trustworthy, are likely to provide accurate and quality information (Metzger, 2007; Rieh, 2002; Rieh & Belkin, 1998). As discussed in the typology, when rendering source-based evaluations of texts’ trustworthiness, students consider author expertise and benevolence (Stadtler & Bromme, 2014). The differences in conceptualizations of source evaluation identified by Stadtler and Bromme (2014) vis-à-vis Petty and Cacioppo (1986) may, in part, be attributable to differences in domain. Evaluations of source, as described by Stadtler and Bromme (2014), have primarily been in scientific domains, about which students may have limited knowledge (Kienhues et al., 2011; Scharrer et al., 2012). As a contrast, Petty and Cacioppo (1986) consider students’ reasoning about a variety of social issues (e.g., university policy around senior exit exams). In this later case, students may be expected to more readily evaluate the content of persuasive arguments on social issues, about which they may be expected to have some thoughts (e.g., death penalty, Lord et al., 1979; affirmative action, gun control, Taber & Lodge, 2006). As this chapter examines how students evaluate inaccurate information in education, we adopt the perspective on source evaluation introduced in the persuasion literature (Chaiken, 1980, 1987; Petty & Cacioppo, 1984, 1986). That is, we consider that students may quickly evaluate sources’ trustworthiness when engaged in heuristic or superficial processing. At the same time, we are particularly interested in students’ perception of text content, when it is presented by a seeming reliable source. Experiential Factors in Educational Contexts In discussing the role of author expertise in message credibility, Stadtler and Bromme (2014) specifically consider the nature of scientific knowledge as not easily verifiable due to students’ limited knowledge and limited access to the resources necessary for scientific investigation. Education seems to suffer from the opposite problem. Through our collective

38    A. LIST and L. D. RUBENSTEIN

and prolonged experiences in schools, both as students, and potentially, as parents or teachers, there emerges an illusion of understanding within education. Because of this experience, everyone may consider themselves experts with insight on how to make education more effective. This illusion may allow educational quackery and pseudo-scientific-based inaccuracies, however well-intentioned, to abound. Such quackery is reflected in a lack of expert knowledge found in some approaches to teacher professional development and evaluation (Frase & Streshly, 1994; Garet, Porter, Desimone, Birman, & Yoon, 2001) and in the rise of “brain games” and other educational technology, purporting to improve cognition, with limited basis in research (Girard, Ecalle, & Magnan, 2013; Simons et al., 2016; Zickefoose, Hux, Brown, & Wulf, 2013). Further, academic achievement is associated with a variety of strong positive and negative emotions, attitudes, and beliefs (Pekrun, 2006; Wang & Eccles, 2012; Weiner, Russell, & Lerman, 1979). Among the most durable inaccuracies persistent in education is that there are a variety of learning styles that account for the diversity in academic achievement persistent in schools (Curry, 1990; Pashler et al., 2008; Riener & Willingham, 2010). While we may wonder about what makes such perspectives on education appealing, when faced with the reality of the negative socioemotional outcomes experienced by students struggling in school (Anderson, Whipple, & Jimerson, 2003; Roeser, Eccles, & Sameroff, 2000; Turner, Husman, & Schallert, 2002), it should come as no surprise that theories emphasizing more holistic, yet individualized aspects of academic achievement have been found to have wide appeal. Within academic contexts, inaccuracies, filtered through individuals’ personal lenses of academic success or failure, may not only appeal but may become particularly durable and resistant to change. In this chapter and in the LAM, we taxonomize the message factors and reader characteristics that impact students’ adoption of persuasive content, particularly under instances of low task involvement (i.e., heuristic or systematic processing) and seeming source trustworthiness. We considered such conditions to reflect circumstances under which inaccuracies may be particularly likely to emerge and be assumed. Moreover, we sought to demonstrate that similar factors that result in students’ adoption of persuasive information also impact students’ adoption of inaccurate information. INTRODUCING THE LIKELIHOOD OF ADOPTION MODEL The LAM is an initial taxonomy specifying the potential message and individual difference factors that, particularly at their intersection, may contribute to students’ adoption of educational inaccuracies (see Figure 3.2). The LAM was developed based on factors in the persuasion literature identified

Understanding Susceptibility to Educational Inaccuracies    39

as contributing to the evaluation and adoption of persuasive information, when students engage in heuristic or peripheral processing. In introducing the LAM, we demonstrate that these factors contribute not only to students’ assimilation of persuasive content but also to their adoption of inaccurate information, in the field of education. Specifically, the LAM presents factors that influence information adoption when learners are engaged in superficial or low-level processing, while considering the task they are completing to be of low value or importance. Chaiken (1980) suggests that, indeed, this may describe the predominant nature of students’ information processing. Moreover, due to its grounding in the literature on persuasion, the LAM specifically considers those situations where students receive information from a seemingly trustworthy or credible sources and, therefore, may not be able to readily dismiss information based only on source features like bias (Kammerer & Gerjets, 2012). Indeed, students may be expected to generally consider sources to be trustworthy, unless obvious cues of source disrepute or explicit instructions to scrutinize information (Brem, Russell, & Weems, 2001; Connor-Greene & Greene, 2002; Kiili, Laurinen, & Marttunen, 2008; List et al., 2017). As such, under conditions of low task involvement and presumed source credibility, a variety of individual and message characteristics may be expected to influence students’ adoption of accurate and inaccurate information. In the LAM, we consider adoption to refer to students’ positive evaluation of the credibility and plausibility of information and their ready assimilation of said information into existing knowledge structures. In terms of individual characteristics, low or moderate prior knowledge (Alexander, Murphy, Buehl, & Sperl, 1998; Kuklinski, Quirk, Jerit, Schwieder, & Rich, 2000), information consistent attitudes or beliefs (Kumkale, Albarracín, & Seignourel, 2010; Lord et al., 1979; Olson & Zanna, 1993; Petty & Cacioppo, 2018; Taber & Lodge, 2006), and low need for cognition (Cacioppo et al., 1983; Haddock, Maio, Arnold, & Huskinson, 2008; Haugtvedt & Petty, 1992) have been identified as impacting students’ evaluation and adoption of information. In terms of message characteristics, a greater variety of factors have been identified. These primarily refer to students’ superficial evaluations of evidence quantity and quality. In terms of quantity, when heuristically processing information, students prefer more evidence to less (Petty & Cacioppo, 1984; Chaiken, 1980; van Amelsvoort & Schilperoord, 2018). Moreover, students have generally been found to rate strong arguments more favorably than weak arguments (McCrudden & McTigue, 2018; Taber & Lodge, 2006). At the same time, students’ identification of arguments as weak or strong has primarily been examined as based on students’ differential ratings of statistical versus anecdotal information (Allen & Preiss, 2009). At the same time, students have been found to have difficulties differentiating the strength of

40    A. LIST and L. D. RUBENSTEIN

correlational vis-à-vis causal statistical evidence (Hoeken & Hustinx, 2003) and anecdotal evidence that is generalizable vis-à-vis analogical (Hoeken & Hustinx, 2007), at least under conditions of superficial processing. As such, while we may conclude that students prefer statistical to anecdotal evidence, it is important to recognize that this preference likely constitutes a fairly superficial determination, wherein the presence of statistical or quantitative information is used as heuristic to determine information quality. In a similar vein, other heuristic features signaling information quality (e.g., references) may also be used by students to determine information acceptability, independent of such features being accurately or appropriately used. Moreover, while students may prefer statistical evidence to anecdotal information, when evaluating anecdotal evidence, there may likewise be superficial features (e.g., use of direct quotation) that students defer to in adopting information. Collectively, the statistical and anecdotal features that students may attend to when adopting information are termed superficial validity features in the LAM. Beyond evidence of quantity and superficial features of quality, linguistic features may influence students’ information evaluation and assimilation. Such features may include directness, versus hedging, expressed certainty, or emotionality. McGuire (2000) found the use of figurative language or rhetorical tropes (e.g., metaphor, emphasis, paradox, hyperbole) to impact students’ evaluations of message content. More recently, Gandarillas, Briñol, Petty, and Díaz (2018) found that tweets expressed via a single hashtag (i.e., one word), rather than as many words, had a greater effect on attitude change, when students were completing a task of low importance. As such, quantitative and qualitative aspects of language use may impact students’ information adoption under conditions of superficial processing. Likewise, although not examined in the persuasion literature per se, the specificity or detail of information presented may impact students’ likelihood of information adoption (Bell & Loftus, 1988, 1989; Krishnamurthy & Sujan, 1999). Indeed, even when the details included are superficial or irrelevant in nature, more specific or elaborated information may prompt students’ more ready adoption of message content. The remaining message factors impacting students’ information adoption have received relatively limited attention in the literature. Nevertheless, a variety of factors related to message comprehensibility, including coherence and ease of understanding (i.e., vocabulary) as well as conceptual simplicity may further impact the likelihood that information will be adopted (e.g., Chaiken & Eagly, 1976; Eagly, 1974; Hafer, Reynolds, & Obertynski, 1996). While students have been found to prefer information that is easier to understand, comprehensibility has most commonly been manipulated linguistically (e.g., as differences in vocabulary, Goldberg & Carmichael, 2017; Kaakinen, Salonen, Venäläinen, & Hyönä, 2011). In the LAM, we suggest that both linguistic and conceptual comprehensibility

Understanding Susceptibility to Educational Inaccuracies    41

(i.e., simplicity) may promote students’ information adoption. Although, to our knowledge, not previously examined in the literature, simplicity may refer to the conceptual clarity of information or to the presentation of discrete, rather than conditional and interconnected, evidence. Indeed, perhaps reflective of a preference for information simplicity, students tend to favor causal explanations when trying to understand relationships between variables (Slusher & Anderson, 1996; Tobin & Weary, 2008). In summary, the LAM proposes that under certain conditions (i.e., low task value, superficial or heuristic processing on the part of learners, and presumed source credibility), the likelihood that an accurate or inaccurate message will be adopted by learners is contingent on a variety of individual and message-related factors. At the individual level, low prior knowledge, low need for cognition, and existing attitudes that are message consistent increase a message’s likelihood of adoption. In terms of message characteristics, argument quantity, seeming superficial validity (reflected in the presence of statistics or the inclusion of references), linguist features (e.g., definitiveness, specificity, comprehensibility, simplicity, and causality) may make the message more likely to be adopted. Of course, the various individual and message characteristics identified and displayed in Figure 3.2 are differentially implicated any time readers decide whether or not to adopt a message. For instance, readers’ adoption of a message may be determined by its seeming certainty and the number of arguments that it presents as well as by their own lack of knowledge or by the simple and causal nature of the message and students’ limited need for cognition. The particular interactions among individual and message related factors and their relative importance have yet to be definitively identified and are explored in the following study. TESTING THE LIKELIHOOD OF ADOPTION MODEL Although a variety of reader and message factors are identified as increasing the likelihood of information adoption in the LAM, three such factors are investigated in this chapter. Specifically, we look at the role of argument quantity (i.e., one or two reasons) and quality (i.e., statistical or anecdotal evidence) in students’ evaluations of information that either conflicts with their existing beliefs or is unrelated to the beliefs that they hold. We examine quantity and quality factors as these have been most commonly explored in the literature on persuasion. Moreover, we examine these factors when providing students with information that is either true, based on the education evidence available, although inconsistent with students’ existing beliefs, or that is fabricated for the purposes of this study. A within-subjects design was used, wherein students read a series of brief texts introducing various educational strategies. Two types of texts were

42    A. LIST and L. D. RUBENSTEIN Likelihood of Adoption

Individual Characteristics: a. Low prior knowledge b. Existing attitudes c. Low need for cognition

Low Task Importance or Value

Message Characteristics: a. Argument quantity b. Superficial validity features (e.g., statistics, references) c. Linguistic features (e.g., directness, certainty, emotionality) d. Specificity e. Comprehensibility f. Simplicity g. Explanatory causality Presumed Source Credibility

Low-Level or Superficial/ Heuristic Processing

Figure 3.3  Likelihood of adoption model. Note: Model features explored in this study are underlined and italicized.

considered: belief inconsistent and novel. Belief inconsistent texts presented strategies (e.g., rereading) that students tend to believe are effective despite research evidence to the contrary. The goal of belief inconsistent texts was to persuade students that these were ineffective strategies. The novel texts suggested a new, fabricated strategy (e.g., taking notes in different colors) with the goal of convincing students of its efficacy. Belief consistent strategies were not examined in this study as prior research has well-established that students evaluate information in belief consistent ways (McCrudden & Barnes, 2016; Taber & Lodge, 2006). Students read eight texts, four presenting belief inconsistent information and four presenting novel strategies for learning. Moreover, within each belief condition, students read texts varying in the quantity (i.e., one or two reasons) and quality (i.e., anecdotal or statistical) of evidence provided to support strategy effectiveness. According to the LAM, we expected students to rate the texts providing more evidence and statistical evidence as more adoptable (i.e., trustworthy and convincing), when compared to texts providing less, anecdotal evidence. We were, however, unsure about how students would rate the novel strategies introduced vis-a-vis the belief inconsistent strategies described. Nevertheless, based on prior work on attitude biased processing, we expected that students would

Understanding Susceptibility to Educational Inaccuracies    43

rate novel strategies higher, due to a trustworthiness discounting effect associated with the presentation of belief inconsistent information (McCrudden & Barnes, 2016; Taber & Lodge, 2006). Participants Participants were 220 undergraduate students at a large university in the North Eastern United States (Age: M = 20.27, SD = 1.85). Students were recruited from two sections of biochemistry course and represented a variety of majors, primarily in the natural sciences. The sample was 66.51% female (n = 145) and 33.49% (n = 73) male. Methods We selected eight topics reflecting four strategies that students already used and considered to be effective (i.e., rereading, taking notes on computer) and matched to four novel strategies that students had likely not encountered previously (i.e., studying from someone else’s notes, taking notes in different colors). The list of strategies identified was based on strategies that undergraduates had previously reported commonly using to study (Hartwig & Dunlosky, 2012). For each strategy, we composed four texts providing either one or two pieces of anecdotal or statistical evidence for strategy effectiveness (see Table 3.1 for samples). The anecdotal texts provided evidence that was based in a particular individual’s personal experience, while the statistical texts summarized the results of a research study. Texts further provided one personal account or one study description, with a number included (e.g., percent improvement) or two personal accounts or two research study descriptions, with two numbers included. A withinsubjects, full factorial design [2 (belief inconsistent vs. novel) × 2 (simple vs. complex) × 2 (anecdotal vs. statistical)] was used. Students were randomly assigned to 1 of 16 conditions, varying in whether one or two pieces of anecdotal or statistical information were provided about each of four belief inconsistent and four novel strategy topics. Within each condition, each of eight texts were presented in a random order. After reading each text, students were asked to rate its convincingness and trustworthiness on a five-point scale from not at all to very. All texts were introduced as brief newspaper stories describing effective strategies for learning, thereby creating an illusion of credibility, despite no specific author information being provided. Consistent with Chaiken’s model (1980), we expected students to engage in superficial or heuristic information processing, absent any instruction or intervention to help them

44    A. LIST and L. D. RUBENSTEIN TABLE 3.1  Samples of Belief Inconsistent Textsa

a

Condition

Sample Text

Single Reason/ Anecdotal

Rereading, although a study strategy commonly used by many students, is actually not an effective strategy for learning. Kevin Williams, a senior biology major, agrees, “When I studied for tests in high school, I reread the textbook and my notes, over and over, until I understood them and that worked for me. When I got to college, I found that I was spending more and more time rereading information, but I still wasn’t doing well on the exams.” Kevin says that after consulting with a TA for his introductory biology course, he realized that he needed a more active method of studying and started using tools like concept mapping and quizzes to review information. He found much better results.

Multiple Reasons/ Anecdotal

Rereading, although a study strategy commonly used by many students, is actually not an effective strategy for learning. Kevin Williams, a senior biology major, agrees, “When I studied for tests in high school, I reread the textbook and my notes, over and over, until I understood them and that worked for me. When I got to college, I found that I was spending more and more time rereading information, but I still wasn’t doing well on the exams.” Karen Hays, a teaching assistant for biology, explains, “A lot of students just reread information when studying, but this leaves students with the same confusions and misconceptions that they had initially.  Students would instead benefit from asking questions or looking at the information in a different way.” Kevin Williams agrees, “Once I started studying in a more active way, instead of just rereading, my performance really improved.”

Single Reason/ Statistical

Rereading, although a study strategy commonly used by many students, is actually not an effective strategy for learning. Researchers at McGill University conducted a study with introductory biology students.  Students attended lectures, and then, were asked to study for a chapter exam. Students were randomly assigned to reread the chapter, write a summary of the chapter, or to create a concept map of the information in the chapter. Students assigned to the concept mapping and summary conditions performed, on average, 18% better than students only asked to reread. Researchers suggest that this is because students learn best when they reorganize and actively manipulate information, such as when concept mapping, rather than passively rereading the same content over and over.

Multiple Reasons/ Statistical

Rereading, although a study strategy commonly used by many students, is actually not an effective strategy for learning.  Researchers at McGill University conducted a study with introductory biology students. Students were asked to study for a chapter exam and were randomly assigned to reread the chapter, write a summary of the chapter, or to create a concept map of the chapter. Students assigned to the concept mapping and summary conditions performed, on average, 18% better than students only asked to reread. Researchers suggest that this is because students benefited from using more active study strategies, rather than passively rereading the same content over and over. A similar study, from researchers at the University of Michigan, tracked students as they prepared for an exam. They found the students who used rereading as their primary study strategy scored 10% lower than students who used different strategies for studying. Rereading is time consuming and takes time away from students using more effective learning strategies or simply getting enough sleep before an exam.

The same format was used to create novel texts.

Understanding Susceptibility to Educational Inaccuracies    45

do otherwise. Students were debriefed after the study and informed that some of the strategies that they read about may not be effective for learning. Moreover, students were provided with resources about what strategies may actually be effective in improving academic performance. Results Two three-way repeated measures ANOVAs were run examining differences in students’ convincingness and trustworthiness ratings according to texts’ belief consistency (i.e., belief inconsistent vs. novel) and the quantity (i.e., one vs. two reasons), and the quality (i.e., statistical vs. anecdotal) of evidence provided. The repeated measures ANOVA for students’ ratings of texts’ convincingness was significant for each target main effect, including belief consistency [F(1, 219) = 18.55, p