Duped Read online

  Duped

  Duped

  Truth-Default Theory and the Social Science of Lying and Deception

  Timothy R. Levine

  The University of Alabama Press

  Tuscaloosa

  The University of Alabama Press

  Tuscaloosa, Alabama 35487-0380

  uapress.ua.edu

  Copyright © 2020 by the University of Alabama Press

  All rights reserved.

  Inquiries about reproducing material from this work should be addressed to the University of Alabama Press.

  Typeface: Scala and Scala Sans

  Cover image: Woodcut style expressionist image of the Greek Trojan Horse © Jeffrey Thompson at www.123RF.com

  Cover design: David Nees

  Cataloging-in-Publication data is available from the Library of Congress.

  ISBN: 978-0-8173-2041-6 (cloth)

  ISBN: 978-0-8173-5968-3 (paper)

  E-ISBN: 978-0-8173-9271-0

  Contents

  List of Illustrations

  Preface

  Acknowledgments

  List of Studies and Experiments

  PART I: THE SOCIAL SCIENCE OF DECEPTION

  1. The Science of Deception

  2. Cues

  3. Deception Detection Accuracy

  4. Rivals

  5. Critiquing the Rivals

  PART II: TRUTH-DEFAULT THEORY

  6. Truth-Default Theory Summarized

  7. Defining Deception (Beyond BFLs and Conscious Intent)

  8. Information Manipulation (Beyond BFLs and Conscious Intent, Part 2)

  9. Prevalence

  10. Deception Motives

  11. Truth-Bias and Truth-Default

  12. The Veracity Effect and Truth–Lie Base-Rates

  13. Explaining Slightly-Better-than-Chance Accuracy

  14. Improving Accuracy

  15. The TDT Perspective

  Notes

  Bibliography

  Index

  Illustrations

  Figures

  1.1 Accuracy in deception detection experiments prior to 2006

  2.1 The relationship between the number of times a cue has been studied (k) and its cumulative effect size (d) weighted by sample size

  3.1 McCornack and Parks’s model of relational closeness and deception detection

  3.2 Variability in judgments of honesty and deception

  9.1 The distribution of the numbers of lies per day in the United States

  9.2 The distribution of the numbers of lies per day in the United States separated by face-to-face and mediated communication

  9.3 The distribution of the numbers of lies per day in the United States separated by target person

  9.4 The distribution of big lies and small lies in the UK

  9.5 Curve fitting results of the US and UK lie prevalence data

  12.1 The relationship between truth-bias and the veracity effect

  12.2 Predicted and actually observed accuracy as a function of truth-lie base-rate in TDT experiment eighteen

  13.1 Accuracy in deception detection experiments prior to 2006

  13.2 Stem-and-leaf plots of judge and sender accuracy in TDT experiment twenty-seven

  13.3 Stem-and-leaf plots of judge and sender accuracy-transparency and believability in experiment twenty-eight

  13.4 Senders matched and mismatched on demeanor and actual honesty

  Tables

  2.1 Cues that affect truth-lie judgments

  2.2 Validity of deception cues in signaling actual deception

  2.3 Associations between cues and actual lying

  3.1 Slightly-better-than-chance deception detection accuracy

  8.1 Initial test of IMT

  8.2 Replication of initial IMT findings

  10.1 Truth-bias and accuracy results of TDT experiments ten, eleven, and twelve

  11.1 Results of TDT experiments thirteen and fourteen

  11.2 Truth-bias in my research

  12.1 Results of TDT experiments thirteen and fourteen revisited

  12.2 “The veracity effect” in my research

  12.3 Using the Park–Levine Model to forecast accuracy at different base-rates using the Leery Lover results from TDT experiment thirteen

  12.4 TDT experiment eighteen results: Accuracy is a predictable linear function of base-rates

  14.1 Results of TDT content-in-context experiments thirty-nine through forty-four

  14.2 Results of TDT experiment forty-five

  14.3 Question sets in the cheating experiments

  14.4 Improving accuracy using different question sets in the cheating interviews

  14.5 Improving accuracy using different question sets in experiments forty-nine, fifty, and fifty-one

  Preface

  DURING THE 2016 US PRESIDENTIAL election, fact-checking revealed that only one out of every six of the winning candidate’s statements was rated as honest or mostly honest. Yet fact-checking apparently held little sway with millions of voters who saw the candidate whose statements were more factually aligned as the less trustworthy of the two. How could this be? Despite the postelection soul searching of political pundits, this pattern is far from unique. Across a host of issues such as the advent of fake news, climate-science denial, and Bernie Madoff’s appeal to investors, people can be astonishingly gullible. There are people who come off as authentic and sincere even when the facts discredit them. People fall victim to conspiracy theories and economic scams that should be dismissed as obviously ludicrous.

  Many people, especially academics, see widespread gullibility as stemming from an unengaged and ill-informed public: If only people were better educated, surely they would be better at correctly distinguishing fact from fiction. Yet simple fact-checking coupled with public-education campaigns is unlikely to provide adequate solutions. Why? Because day in and day out, we spend our lives within a mind-set that can be characterized as a “truth-default.” We uncritically accept virtually all of the communication messages we receive as “honest.” Think about it: how many tweets, posts, articles, texts, e-mails, phone calls, and spoken declarative sentences do you receive each day? Now ask yourself: How many of those do you question in terms of honesty? Chances are, the answer is near zero. This is a near-universal human tendency. We all are perceptually blind to deception. We are hardwired to be duped. The question is, can anything be done to militate against our vulnerability to deception without further eroding the trust in people and social institutions that we so desperately need in civil society?

  But there is a second critical point. Even though we have a strong tendency toward unquestioning belief, sometimes we do suspect deception. There are situations in which we abandon our truth-default. But, even when people are on guard for deception, folk wisdom about deception leads to predictable errors in judgment. In the most recent presidential campaign, blurting out politically incorrect statements was understood by many Americans as evidence of authenticity. Appearing confident was interpreted as a sign of honesty. Alternatively, speaking like a politician with carefully chosen words signaled that the candidate was not to be trusted. Rather than questioning whether the candidate’s statements actually aligned with known facts, assessments of honesty and sincerity were based on the candidate’s demeanor. The problem is, demeanor is highly misleading. Appearing honest and being honest are usually unrelated.

  My objectives here are ambitious and radical. I want to start a revolution. I seek to overthrow existing deception theory and provide a new, coherent, and data-consistent approach to understanding deception and deception detection. For more than twenty-five years, I have seen a need for a new theory of deception and deception detection. Ekman’s idea of leakage was hugely influential, but the deficiencies were apparent almost immediately. His focus shif
ted over time from the leakage hierarchy to a focus on the face and microexpressions. But my read of the ensuing literature reveals more excuses for why the data do not seem to support his theory than solid, replicated, affirmative scientific support. Interpersonal deception theory is even less viable. It is logically incoherent, and I knew it to be empirically false four years before it was eventually published. The new cognitive load approach in criminal and legal psychology does not seem to be the path forward either, for the theoretical reasons identified by Steve McCornack, as well as weak, inconsistent, and just plain odd empirical findings. The need is clear. Existing theory does not cut it. A new perspective is needed.

  Duped: Truth-Default Theory and the Social Science of Lying and Deception tells the story of my program of research culminating in my new theory of deception: truth-default theory (TDT). Approximately sixty original studies and experiments are summarized and woven together within the TDT framework. I detail where the ideas came from, how ideas were tested, and how the findings combine to produce a coherent new understanding of human deception and deception detection.

  The story begins in 1989, when I coauthored my first deception detection experiment as a graduate student. The experiment brought college-student dating couples into the lab, and my professor and I looked at how the closeness of the communicators’ relationship and how prompting suspicion affected our research subjects’ ability to tell whether their partners were lying. The big finding was that even when we tried to make them highly suspicious, our subjects still tended to believe their partners regardless of their partners’ actual honesty. This finding was called truth-bias, and it has turned out to be a very robust finding. Since then I have collected data in countries around the world and recruited a wide variety of research subjects, including college students, university professors, police detectives, customs agents, and professional spy catchers. Truth-bias has been a constant finding. I have never found people to be otherwise. Over time I have pieced together coherent understandings of truth-bias, how best to catch lies, and the interplay between the two.

  TDT proposes that the content of incoming communication is usually uncritically accepted as true, and most of the time this is a good thing for us. One of the most surprising new insights is that truth-bias and truth-default work well for us. I argue that the tendency to believe others is an adaptive product of human evolution that enables efficient communication and social coordination. The truth-default allows humans to function socially. Further, because most deception is enacted by a few prolific liars, the so-called truth-bias is not really a bias after all. Passive belief makes us right most of the time. The catch is that it also makes us vulnerable to occasional deceit.

  Importantly, TDT also challenges current social-scientific and folk views of deception that prioritize nonverbal and linguistic behavior as the keys to lie detection. According to TDT, the path to improved human lie detection involves listening to what is said, rather than to how it is said. As previously mentioned, the recent US election provides an excellent example of a situation where confidence and belligerence were decoded as authenticity and were often mistaken for honesty. More broadly, research shows that using evidence and skilled question asking produces much better outcomes than passive observation of deception cues like gaze aversion, facial expressions, body language, pronoun usage, or decontextualized counts of details.

  My research on lie detection and truth-bias has produced many provocative new findings over the years. For example, we have uncovered what makes some people more believable than others (a believability quotient) and have discovered several ways to improve lie-detection accuracy. I directed a team of researchers (funded by the FBI and involving agents from the NSA) that produced the highest lie-detection accuracy ever published. Truth-default theory weaves together all these findings under a common framework. The evidence is organized and presented all in one place.

  This book’s perspective is best described as quantitative, experimental, multidisciplinary social science. My academic home discipline is communication. Consequently, deception and deception detection are first and foremost understood as communicative processes. The focus is more social and interpersonal than intrapsychic. Theoretically, TDT integrates ideas from a variety of academic disciplines such as evolutionary biology (Robert Trivers), social psychology (Dan Gilbert), and the philosophy of language (Paul Grice). Structurally, the book is modeled on classics in experimental social psychology such as Latane and Darley’s (1970) The Unresponsive Bystander, presenting a logical series of original experiments.

  The research subjects in TDT experiments range from college-student dating couples to nationally representative samples of US adults to elite NSA agents. Data collections span five continents. Every major TDT claim is scientifically tested and replicated. Small chunks of TDT ideas and results have been appearing in peer-reviewed academic journal articles for more than two decades. The book shows how all those findings and ideas fit together into a coherent package. The book also showcases for students, young professors, and social science aficionados what can be achieved with long, sustained, ambitious, programmatic, multimethod research. The applications are diverse, ranging from deception in romantic relationships to catching terrorists to criminal interrogation.

  My approach might be described as abductive science. I don’t see what I do as either dust bowl empiricism or as exclusively hypothetico-deductive theory testing. The theory building presented in this book is not of the abstract, armchair, speculative sort. The propositions are all data based, and the explanations were articulated so as to offer a coherent account of the existing scientific data. I did not seek to publish my theory until I had original research to support and replicate every major claim. I have that evidence now, and it is presented in the pages that follow.

  But my intent is not just post hoc explanation. Good theory must also be generative. It needs to lead to new predictions that no one would think to make absent the theory. In line with Imre Lakatos, I want to propose a theory that is out in front of the data, not always chasing from behind to try and catch up.

  A final feature of my theory is that it is modular. Truth-default theory is a collection of quasi-independent minitheories, models, or effects that are joined by an overarching logic. The parts can stand on their own, but they also fit together to provide a bigger picture.

  Stylistically, the text shifts between first-person storytelling, objective scientific narration, and editorializing. The book provides an engaging read targeted to a diverse audience. I strove to be accessible to the novice while also providing valuable insights for even the most sophisticated and informed reader. I aim to provide an engaging, fun, interesting read without sacrificing scholarly rigor. The result is a book that can be read and appreciated at different levels by different audiences.

  Acknowledgments

  THERE ARE SEVERAL INDIVIDUALS WITHOUT whom this book would not have happened, and many of the ideas presented here are not all my own. First, there is my good friend Steve McCornack, who, to my knowledge, first coined the term “truth-bias.” Perhaps no other idea has played a more prominent role in my thinking. Without Steve, it is unlikely I would have ever done a deception experiment. As I explain in chapter 1, I got my start in deception research as Steve’s lab assistant. My first several published articles on deception were all collaborations with Steve. His ideas regarding truth-bias, suspicion, probing, information manipulation, and the problems with cognitive load all influenced my thinking in important ways. I am grateful for Steve’s friendship as well as his intellectual contributions.

  Another absolutely critical influence was Hee Sun Park. Besides being my spouse, she came up with the ideas for the veracity effect, the Park–Levine probability model of base-rate effects, and the “How People Really Detect Lies” study. Each of these ideas is a key part of truth-default theory (TDT). Without these critical pieces of the puzzle, I would not have a coherent theory of deception.

  A third key con
tributor was J. Pete Blair. Prior to earning his PhD, Pete was a professional investigator and interviewer-interrogator. He has brought a more applied flavor to my research, and many of the ideas regarding improving accuracy (e.g., content in context, expertise, question effects) are, at least in part, Pete’s. Pete has become a highly valued collaborator and a good friend.

  Three of my former graduate students deserve special mention. Ms. Rachel Kim was my chief lab assistant and collaborator on the creation of the NSF tapes, as well as experiments on base-rates, suspicion, deception motives, and projected motives. I owe Rachel much, and it was a pleasure having her on the team. When Rachel left MSU, I was exceptionally fortunate to have David Clare step into the role as my chief lab assistant. David was there for many of the more recent data collections (e.g., demeanor, experts, interactive base-rates, and the FBI expert experiment), and he was a fabulously reliable research assistant and a great student. David’s preliminary PhD paper provides key evidence for the central premise of the theory. Kim Serota played an important role in the “few prolific liars” program of research, the demeanor studies, some of the base-rate studies, and the Lie to Me experiment. Amongst many assets, Kim has a real gift for the visual depiction of data. Kim created or formatted several of the figures presented in the book. He continues to be a valued friend and coauthor and has taken the lead in the lie prevalence research. His restaurant recommendations are spot-on too.