Accuracy of Deception Judgments

Charles F. Bond, Jr., Texas Christian University

and

Bella M. DePaulo, University of California at Santa Barbara

We are grateful to Harris Cooper, Laura Muhlenbruck, Bob Rosenthal, and Jenny Tornqvist for help with this project. We also thank the many deception researchers who answered questions about their work.

Address correspondence to Charles F. Bond, Jr., Department of Psychology, Box 298920, Texas Christian University, Fort Worth, TX 76132 USA or to the e-mail address C.BOND@TCU.EDU.

We analyze the accuracy of deception judgments, synthesizing research results from 206 documents and 24,483 judges. In relevant studies, people attempt to discriminate lies from truths in real time with no special aids or training. In these circumstances, people achieve an average of 54% correct lie/truth judgments, correctly classifying 47% of lies as deceptive and 61% of truths as non-deceptive. Relative to cross-judge differences in accuracy, mean lie/truth discrimination abilities are non-trivial, with a mean accuracy d of roughly .40. This produces an effect that is at roughly the 60^th percentile in size, relative to others that have been meta-analyzed by social psychologists. Alternative indices of lie/truth discrimination accuracy correlate highly with percentage correct, and rates of lie detection vary little from study to study. Our meta-analyses reveal that people are more accurate in judging audible than visible lies, that people appear deceptive when motivated to be believed, and individuals regard their interaction partners as honest. We propose that people judge others’ deceptions more harshly than their own, and that this double standard in evaluating deceit can explain much of the accumulated literature.

Deception entered Western thought in a telling guise when the author of Genesis placed a serpent in the Garden of Eden. By lying, the serpent enticed Eve into committing the original sin. Thus deception was enshrined as the ultimate source of evil.

Lying has always posed a moral problem. Aristotle wrote that “falsehood is in itself mean and culpable”; St. Augustine believed that every lie is a sin; and Kant regarded truthfulness as an “unconditional duty which holds in all circumstances.” Others take a more permissive stance. Aquinas countenanced lies told in the service of virtue, while Machiavelli extolled deceit in the service of self. For background on these ethical matters and a contemporary position, see Bok (1989).

Having been a moral issue for millenia, deception came also to be viewed as a legal challenge. Since Diogenes, many had suspected that lying was commonplace and could have pernicious influences on human affairs. The chore of truth finding fell to the legal system, and procedures for lie detection were devised. Over the centuries, authorities employed a number of unsavory means to extract legal “truths” (Trovillo, 1939). Modern sensibilities inspired some of the current techniques: religious oaths, cross-examinations, threats of incarceration. Technological developments have had an impact too. The polygraph, the psychological stress evaluator, brain fingerprints, EEGs – these have been promoted for their ability to divine deception. Yet in the first decade of the 21^st century, American jurisprudence entrusts lie detection to ordinary citizens. U.S. courts bar technological aids to lie detection and deception experts too. Witnesses must appear in person before jurors who are the “sole judges” of the witness’s believability. American jurors are instructed to judge the person’s truthfulness by considering his or her “demeanor upon the witness stand” and “manner of testifying” (Judicial Committee on Model Jury Instructions for the Eighth Circuit, 2002; p. 53). According to an official view, this system of lay judgment solves the legal problem of deception because “lie detecting is what our juries do best” (Fisher, 1997).

A moral problem for millenia and a legal problem for centuries, deception has more recently become a research problem. How successful are people at deceiving others? How likely are they to believe others’ fibs? What accounts for liars’ successes and failures? When and why are people duped? These questions are of moral and legal interest. The ethics of lying would be moot if people were rarely duped. Current legal practices would be called into question if ordinary people could not spot deception when they saw it.

In the current article, we summarize research on 4435 individuals’ attempts to dupe 24,483 others. We offer quantitative measures of deceptive success and identify conditions under which people are more and less gullible. As a background for our statistical synopses, we summarize some earlier characterizations of deception, sketch a new framework for understanding this subject, and consider earlier research reviews.

Characterizations of Deception

“No mortal can keep a secret. If his lips are silent, he chatters with his finger-tips; betrayal oozes out of him at every pore.” Freud (1905)

With this quotation, Ekman and Friesen (1969) open a pioneering article on the psychology of deception. Where Freud had analyzed verbal slips to self-deception, Ekman and Friesen describe nonverbal signs of individuals’ attempts to deceive one another. These authors discuss lies that involve high stakes and strong emotion. In their view, liars face challenges. They must guard against nonverbal “leakage” of feelings they are trying to conceal and must hide their own affective reactions to the act of lying, such as guilt, anxiety, and shame. People find it especially difficult to lie in certain situations: when the possibility of deception is salient to both actor and target, when the target can focus on detecting deception without concern for his/her own behavior, and when the actor and target have antagonistic interests (the actor wishing to perpetrate deceit and the target to uncover it).

Ekman and Friesen (1969) offer a theory about the anatomical locus of nonverbal deception cues. They predict that people are most likely to show deception in the legs and feet, less likely to show it in the hands, and least likely to show deception in the face. These predictions follow from a communicative analysis: relative to the face, the feet and legs have a weak sending capacity, generate little internal feedback, and occasion few reactions from others. Thus, people have more ability and motivation to control the face than the feet and legs. By this logic, people have intermediate ability and motivation to control the hands.

Thirty-two years later, Ekman (2001) emphasizes the ambiguity of nonverbal deception cues. There being no foolproof sign of deceit, many inferences of deception are mistaken. In trying to spot lies, people must avoid untoward influences of their own suspicions as well as misinterpretations of others’ idiosyncrasies. Ekman attributes failures at lie detection to many factors: poor evolutionary preparation, socialization to overlook lies, the psychological benefits of trust, and inadequate feedback from errors.

Ekman’s work has been influential. It has encouraged nonverbal analyses that aim to expose deceit. Inspired by Ekman’s early work, Miller and Stiff (1993) enumerate cues to deception and cues to judgments of deception, then attribute failures at spotting deception to differences in the two sets of cues. Pursuing Ekman’s emphasis on high-stakes deceit, forensic psychologist Vrij (2000) discusses the implications of experimental findings for professional lie catchers.

Buller and Burgoon (1996) propose a theory for face-to-face deceptive interactions. In order to dupe others, people must craft a verbal deception, bolster it with ancillary strategic messages, and suppress discrediting behaviors. Meanwhile, the targets of face-to-face deceit must manage behavioral signs of suspicion. Burgoon and Buller (1996) trace the unfolding of deceptive exchanges over time. Theoretically, receivers are more likely to perceive a person as truthful if they are interacting with that person – rather than seeing the person on videotape. Theoretically, deceivers should be more likely to engage in strategic activity and less likely to engage in non-strategic activity in interactive contexts. In interactive contexts, deceivers react to signs of suspicion, and targets react to indications that their suspicions have been surmised.

Critical of Buller and Burgoon’s approach (DePaulo, Ansfield, & Bell, 1996), DePaulo and colleagues favor a self-presentational perspective on deception (DePaulo, 1992; DePaulo, Kashy, Kirkendol, Wyer, & Epstein, 1996; DePaulo, Lindsay, Malone, Muhlenbruck, Charlton, & Cooper, 2003). In this view, lying is a part of everyday life. People tell lies to avoid embarrassment and make positive impressions. They fib on the spur of the moment without compunction, telling polite lies of little consequence. Some everyday lies are scripted and require less cognitive effort than meticulously truthful statements. Occasionally, people tell lies to hide transgressions. Most of these serious lies involve a self-presentational stake: the liar’s reputation. In this view, the signs of deception are subtle, and social norms encourage people to accept others’ representations at face value.

A Double Standard

Having reviewed some earlier characterizations of deceit, let us offer a new framework for understanding this subject. We believe that there is a double standard in evaluating deception.

Our framework begins by noting that people regard truth telling as unexceptional. They accept most statements at face value, rarely inquiring into the authenticity of what they hear. People come to regard an assertion as truthful only after entertaining the possibility that it was deceptive. Then they see truthfulness as a virtue. People are proud of themselves for speaking the truth. People who are told the truth praise truth tellers, and psychologists praise them too. No doubt, there are limits to the morality of truthfulness. Truths are seen as most virtuous when they oppose the truth teller’s interest. Occasionally, people volunteer truthful observations that hurt others, and these are ethically dubious. In most cases, however, truth telling is non-problematic. Problems arise not from truth telling but from deception.

There are two perspectives on deception. One is the perspective that people hold when they themselves are lying; a second is the perspective they bring to others’ lies (Gordon & Miller, 2000). As deceivers, people are practical. They accommodate perceived needs by lying. Of course, deceivers rarely regard their own falsehoods as lies but as something more innocuous. People may lie in the interest of impression management (DePaulo et al, 2003) or for more tangible ends. They exaggerate, minimize, and omit. They give misleading answers to questions. Regarding half-truths and self-editing as necessities of social life, deceivers see deception as similar to these sanctioned practices. Animated by momentary exigencies, offered in passing, lies occasion little anxiety, guilt, or shame (DePaulo, Kashy, Kirkendol, Wyer, & Epstein, 1996). They are easy to rationalize. Yes, deception may demand construction of a convincing line and enactment of appropriate demeanor. Most strategic communications do. To the liar, there is nothing exceptional about lying.

If pragmatic about their own deceptions, people become moralistic when they consider others’ lies (Saxe, 1991). Then deception is wrong and reflects negatively on the deceiver. Indeed, people view duplicity as one of the gravest moral failings. In their ratings of 555 personality trait terms, college students rate as least desirable the trait of being a liar (Anderson, 1968). Social logic assumes that honest people always act honestly (Reeder, 1993); thus, to label a statement a lie is to imply that the person who made that statement is a liar (O’Sullivan, 2003). This is a serious accusation. People have a prescriptive stereotype of the liar -- stricken with shame, wracked by the threat of exposure, liars leak signs of their inner torment. They fidget, avoid eye contact, and can scarcely bring themselves to speak – a worldwide stereotype holds (Global Deception Research Team, 2004). The stereotypic reasons for lying are nefarious too – terrorists lying to further their murderous plots, charlatans scheming to bilk the innocent, husbands cheating on their faithful wives. As old as the Garden of Eden, this moralistic perspective on deceit underlies current psychological thought.

Let us sketch a few implications of the double standard in evaluating deception. People hold a stereotype of the liar – as tormented, anxious, and conscience-stricken. Perceivers draw on this stereotype when considering a target’s veracity. Targets who most resemble the stereotype are most likely to be regarded as liars; those who least resemble it are most likely to be believed. Factors that influence a person’s likelihood of appearing tormented, anxious, or conscience-stricken should affect the person’s judged truthfulness. One such factor would, we suspect, be the stakes surrounding a speaker’s credibility. Most lies are little. When telling white lies of the sort often studied by researchers, people have no reason to appear tormented. Thus, they should often be judged truthful. Occasionally, the stakes of being believed are big. When facing huge stakes, people who ruminate over their credibility may come to match the liar stereotype. Then they would be judged deceptive, even if they were telling the truth.

In the current article, we consider veracity judgments in light of the double standard for evaluating deception. We do not confine attention to momentous lies or evil deceit of the sort most would associate with others’ deception. Rather, we consider all falsehoods that have been studied and hope to use the accumulated literature to learn about people’s successes in engineering various forms of deception. We will credit people for their successes at perpetrating deceit, while noting some unintended consequences of observers’ moralistic stance.

Research on Detection Accuracy

To understand deception, researchers conduct experiments. They arrange for people to lie and tell the truth, and for others to judge the veracity of the resulting statements. For convenience, we will be calling the peoples who lie in these experiments senders, the truthful and deceptive statements messages, and the people who judge these messages receivers. We are interested in receivers’ accuracy in judging senders’ veracity. We will not be reviewing all attempts at lie detection. Rather, we confine attention to receivers who must judge deceit without the aid of polygraphs, fMRIs, or other physiological devices; receivers who judge deception from a brief encounter with an unfamiliar sender in real time. These deception judgments are based on verbal content and the liar’s behavior. Here we review earlier summaries of this research.

Often, lie detection abilities are expressed on a familiar scale: percentage correct. In relevant experiments, receivers classify messages as either lies or truths; hence across messages, the percentage of messages a receiver correctly classifies can be used as an index of his/her detection ability. Ordinarily, half of the messages a receiver encounters are truths, and half are lies; hence by guessing a receiver could expect to achieve 50% correct classifications.

Kraut (1980) offered a statistical summary of results from 10 such experiments. Finding a mean accuracy rate of 57%, Kraut concluded that “the accuracy of human lie detectors is low.” In a summary of 39 studies published after 1980, Vrij (2000) replicated Kraut’s finding, discovering that receivers of more recent research achieve an average of 56.6% accuracy. Along with narrative reviews of the research literature, these statistical summaries have inspired a consensus -- “it is considered virtually axiomatic . . . that individuals are at best inaccurate at deception detection” (Hubbell, Mitchell, & Gee, 2001).

Although it may be “virtually axiomatic” that people are poor at detecting deception, we are reluctant to accept this conclusion on the basis of existing work. We agree that in 50 (or so) pertinent studies people achieve 50-60% correct when classifying messages as lies or truths. However, meta-analyses of percentage correct omit evidence relevant to ascertaining the accuracy of deception judgments. In the omitted experiments, receivers rate the veracity of lies and truths on multi-point rating scales. There accuracy is not gauged in terms of percentage correct – but as a difference between the rated veracity of truths vs. the rated veracity of lies.

Three statistical summaries of lie detection accuracy have incorporated rated-veracity results. They quantify the degree to which lies can be discriminated from truths by a standardized mean difference (d): the mean difference between obtained and chance accuracy in a study divided by a standard deviation from that study. Applying this metric to the results of 16 early studies, DePaulo, Zuckerman, and Rosenthal (1980) calculated a median d of .86 standard deviations. Twenty years later, Mattson, Allen, Ryan, and Miller (2000) found an average difference between the judged veracity of lies and truths of d = 1.07 standard deviations in 7 studies of organizational deception.
Assessing the accuracy of deception judgments in various media, Zuckerman, DePaulo, and Rosenthal (1981) found that receivers who have access to speech regard lies as less credible than truths with a mean d = 1.14.

How strong are the levels of lie detection found in these rated-veracity reviews? To answer this question, it may be helpful to consider results found in other lines of research. From a large-scale compilation, Richard, Bond, and Stokes-Zoota (2003) developed empirical guidelines for evaluating effect sizes. These scholars describe a d of .20 as small, a d of .40 as medium, and a d of .60 as large because these values would be larger than the average standardized mean differences found in 30%, 50%, and 75% of 474 social psychological research literatures the scholars reviewed. Compared to these reference values, people would seem to have a strong ability to detect deception. The median d of .86 standard deviations found by DePaulo, Zuckerman, and Rosenthal (1980) would place lie detection accuracy at roughly the 85th percentile in size, relative to 474 social psychological effects (Richard, Bond, & Stokes-Zoota, 2003). The ability to detect audible lies (mean d = 1.14 standard deviations: Zuckerman, DePaulo, and Rosenthal, 1981) is even better -- ranking at the 95th percentile of 474 social psychological effects.

While amassing evidence on receivers’ accuracy in discriminating lies from truths, scholars have been interested in a more general judgmental tendency – a bias to perceive messages as truthful. By virtue of the bias, truthful messages are more often detected than deceptive messages. Summarizing 15 studies, Zuckerman, DePaulo, and Rosenthal (1981) express this accuracy difference in standard deviation units and find a mean d = .86. Vrij (2000) summarizes 9 percentage-correct studies to find a strong truth bias – a mean of 61.5% truth judgments, 67% accuracy for truths and 44% accuracy for lies.

The Present Review

Given the moral and legal significance of deception, it is important to know how often people are duped. Although previous work provides some hints about people’s success in deceiving others, the work has limitations. The largest review to date is based on 39 research studies. Here we summarize evidence from 206 studies. Some of the previous reviews express the accuracy of deception judgments as a standardized mean difference while others gauge accuracy in terms of percentage correct. Each of these measures has limitations. Standardized mean differences can be hard to interpret (Bond, Wiitala, & Richard, 2003), and meta-analyses of percentage correct cannot include results on rating scale judgments of deception.

Here we assess the accuracy of deception judgments in terms of percentage correct, the standardized mean difference, and with some indices that statisticians favor – the log odds ratio and d’ (Swets, 1996). Perhaps the pattern of results across various measures of accuracy can help resolve a tension in earlier meta-analytic results – between the strong detection abilities implied by standardized results and an “axiom” of inaccurate lie detection in percentage correct (Hubbell, Mitchell, & Gee, 2001).

Some have thought that detection performances vary “only slightly” across situations (Kalbfleisch, 1990); while others have concluded that performance variance across situations is “considerable” (Miller & Stiff, 1993). Here we provide the first test to date of the possibility that there is no variance in detection performances across situations. Assuming that there is such variance, we will provide the first estimates to date of the magnitude of these situational differences. We will also have the opportunity to document the impact of various factors on the accuracy of deception judgments, like the medium in which deception is attempted, the liar’s motivation, and the judge’s expertise. The evidence may have implications for theories of deception, including our double standard framework.

Method

Literature Search Procedures

To locate relevant studies, we conducted computer-based searches of Psychological Abstracts, PsycInfo, PsycLit, Communication Abstracts, Dissertation Abstracts International, WorldCat, and Yahoo through August of 2005 using the keywords deception, deceit, and lie detection; searched the Social Sciences Citation Index for papers that cited key references (e.g., Ekman & Friesen, 1974); examined reference lists from previous reviews (DePaulo et al. 1985a; Zuckerman et al., 1981; Zuckerman & Driver, 1985), and reviewed the references cited in more than 300 articles on the communication of deception from our personal files plus all references cited in every article we found. We sent letters requesting papers to scholars who had published relevant articles.

Criteria for Inclusion and Exclusion of Studies

Our goal was to summarize all English-language reports of original research on the accuracy of judgments of lies and truths available to us prior to September 2005. To be included in this review, a document had to report a measure of accuracy in discriminating lies from truths.

We excluded studies in which individuals judged only lies and those in which individuals judged only truths. We excluded studies in which judges received experimental training or instructions about how to detect deception, studies in which judges received attention-focusing instructions, studies in which senders and receivers knew one another prior to the study, and studies in which individuals could incorporate into their judgments systematic aids to lie detection (e.g., polygraph records, CBCA, or behavior codings from repeated viewings of a videotape). We excluded reports that were not in English, judgments for lies and truths told by senders who were less than 17 years old, as well as judgments made by receivers who were less than 17. We excluded reports in which senders role-played an imagined person in an imagined situation. We also excluded all results on implicit deception judgments (implicit judgments having recently been meta-analyzed by DePaulo et al, 2003), and on judgments of affect (even affects that people were trying to conceal). We uncovered 206 documents that satisfied our inclusion criteria. For a listing of these documents, see Appendix A which is available on-line.

Identifying Independent Samples

Research studies in this literature exhibit two forms of interdependence: sender interdependence and receiver interdependence. Senders are interdependent when the lies and truths told by a given sample of senders are shown to multiple samples of judges. Receivers are interdependent when researchers report multiple measures of lie/truth accuracy for a given sample of judges. The unit of aggregation in the current meta-analysis is the receiver sample. The primary analyses below extract one measure of lie/truth discrimination accuracy from each independent sample of judges – even in those cases where several samples are judging the same lies and truths. For these analyses, our data set consists of 384 independent samples. To assess the impact of moderator variables, we disaggregated receiver samples to reflect within-receiver experimental manipulations.

Variables Coded From Each Report

From each report, we sought information about the following variables: a) number of senders, b) number of receivers, c) percentage correct, d) percentage truth, e) an accuracy standardized mean difference, f) sender motivation, g) receiver motivation, h) sender preparation, i) sender interaction, j) receiver expertise, k) judgment medium, and l) baseline exposure. For our coding of these variables in each of 384 receiver samples, see Appendix B on-line.

Let us explain these variables. The number of senders and number of receivers were coded from each document. From each document that reported results on dichotomous lie-or-truth classifications, we noted percentage correct – more precisely, the unweighted average of the percentage of truthful messages correctly classified and the percentage of deceptive messages correctly classified. Of our 384 receiver samples, 343 judged 50% lies and 50% truths. In these cases, the unweighted average was the overall percentage correct. Whenever authors reported the overall percentage of messages classified as truthful, this percentage truth judgments was coded. From each document that reported results on rating-scale veracity judgments, we noted an accuracy standardized mean difference -- defining d as the mean veracity rating of truths minus the mean veracity rating of lies divided by a standard deviation. As Kalbfleisch (1990) notes, deception researchers’ reporting of standard deviations poses challenges for meta-analysts. Whenever possible, we used as our standard deviation a pooled within-message standard deviation across receivers. In such cases, we would note the variance across receivers in judgments of the veracity of truthful messages and the variance across receivers in judgments of the veracity of deceptive messages, before taking the square root of the average of these two variances. When necessary, we used other standard deviations – for example, the standard deviation across receivers in the difference between the mean rated veracity of truths and the mean rated veracity of lies.

The other variables of interest to us are categorical. People can try to detect lies over various media. Here we coded deception medium by noting whether a given sample of receivers was trying to detect lies over a video medium, an audio medium, an audiovisual medium, or some other medium. We coded sender motivation by noting whether participants had any special motivation to succeed at deception. Our coding of sender preparation reflected whether the senders in a study had any time to prepare their lies and truths. We coded whether or not receivers got a baseline exposure to the sender before making deception judgments.

In some studies, senders are interacting with others as they lie and tell the truth; in other studies, they are not. For purposes of coding sender interaction, we regarded senders as not interacting if when lying they were alone or in the presence of a passive observer. We deemed all other senders to be interacting, and noted whether or not the interaction partner was the receiver (e.g., the person who was judging deception). Most of the receivers in this literature are college students. Others are people whose occupations are thought to give them special expertise at lie detection. We noted this variable of receiver expertise.

We coded the status of the report as published or unpublished. In some instances, the same data are reported in two places – say, a dissertation and a journal article. In such cases, we have listed the more accessible report in the References below. Occasionally, results from a given study are more fully reported in one document than another. Then we used the more complete reporting even if it was from the less accessible document.

Reliability of Coding

For a reliability check, the two authors independently coded 24 of the documents Appendix A. These were selected at random, subject to the restriction that no individual appear as an author on more than two documents. The 24 documents we selected in this manner contribute 46 independent receiver samples to our meta-analysis, and it is on these 46 receiver samples that reliability data are available. The following quantitative variables were checked: number of senders, number of receivers, percentage correct, percentage truth, and accuracy d. Reliabilities on these variables were uniformly high; lowest Pearson’s r = .894 for 10 accuracy ds. We also checked coding of the following categorical variables: sender motivation, receiver motivation, sender preparation, sender interaction, judgment medium, and baseline exposure. For the percentage agreement on each of these variables, see Table 1.

Results

Characteristics of the Literature

We found 206 documents that satisfied our criteria – 133 that were published and 73 that were unpublished. The earliest document was dated 1941, and the latest was published in 2005. Half of these documents were dated 1994 or earlier.

The documents reported results on 24,483 receivers’ deception judgments of 6651 messages offered by 4435 senders. There were 177 independent samples of senders, and 384 independent samples of receivers. One hundred and ten of the sender samples were judged by only a single receiver sample; at the other extreme, one sample of senders was judged by 13 independent receiver samples.

In 277 receiver samples, participants classified messages as lies or truths; in 92 samples, they judged messages on multi-point rating scales; and in 15 samples, receivers made lie-or-truth classifications as well as multi-point ratings. For some other characteristics of this literature, see Table 1. In a typical research study, 41 receivers made judgments of 16 messages – one message offered by each of 16 senders. The typical message lasted 52 seconds. In most cases, the judgment medium was audiovisual, and receivers had no baseline exposure to the sender. Although about 55% of the sender samples had no particular motivation to succeed when lying, over 40% were motivated. Receivers were rarely motivated; barely 12% of the receiver samples had any special incentive to succeed at lie detection. In a little over half of the samples, receivers were judging senders who had had time to prepare their lies; in about 65% of the samples, receivers judged senders who were interacting as they lied. Although only 12% of the receiver samples could claim any occupational expertise in detecting deception, this was nonetheless 2842 experts.

Percentage Correct

In 292 samples, receivers classified messages as lies or truths. From each such sample, we noted the mean percentage correct lie/truth classifications. These are shown on the right side of Figure 1 as a stem-and-leaf display. As can determined from the display, over three-fourths of these means are greater than 50% and less than one in seven is greater than 60%. Across all 292 samples, the unweighted mean percentage correct lie/truth classifications is 53.98%. The highest mean percentage correct attained in any sample is 73%, and the lowest is 31%. Means at the first, second, and third quartile are 50.07%, 53.90% and 58.00%.

Further insight into lie/truth discrimination abilities can be gleaned from Figure 2, which displays the mean percentage correct lie/truth classifications in a study as a function of the total number of judgments on which the mean was based. The latter was determined by multiplying the number of receivers in a sample by the number of judgments each receiver rendered. Note, for example, the right-most point in the plot. This represents the mean lie/truth discrimination accuracy of 54.30% observed by DePaulo and Pfeiffer (1986) in 10,304 dicomotous lie/truth judgments (64 judgments made by each of 161 receivers).

Figure 2 display exhibits a funnel pattern (Light, Singer, & Willett, 1994) with high variability among means based on small numbers of judgments and low variability among means based on large numbers of judgments. This pattern suggests that the studies are estimating a common value and that small sample sizes account for much of the variability toward the left of the plot.

A formal analysis of between-study differences begins by noting that the observed standard deviation in mean percentage correct is only 6.11% (that is, variance = 37.33%). Statistically, results would vary some from study-to-study merely by virtue of different investigators examining different receivers. Random-effects techniques can be used to separate between-study variance due to sampling variability from true variance (Hedges & Vevea, 1998). Using a weighted method of moments technique, we infer that receiver sampling error accounts for 45.29% of the observed between-study variance in mean percentage correct, and that the true standard deviation across studies in mean percentage correct is only 4.52%.

For other analyses of mean percentage correct, we used procedures outlined by Bond, Wiitala, and Richard (2003). These require an estimate of the standard deviation in percentage correct in each study. Whenever a standard deviation was reported (or could be calculated), we used it. Otherwise, we imputed the standard deviation across the receivers in a sample from the binomial distribution, using the mean sample percentage correct as well as the number of judgments made by each receiver in that sample.

These weighted techniques reveal a mean of 53.46% correct lie/truth classifications; 95% confidence interval = 53.31-53.59%. This mean is significantly greater than 50%, t(7994) = 39.78, p < .0001. Between-study variability (though small in size) is greater than would be expected by chance, F_w(283,3658) = 12.61, p < .0001.

Standardized mean differences

Having found that dichotomous lie-or-truth classifications are correct slightly more than half of the time, we next wished to gauge receivers’ ability to distinguish lies from truths on multi-point rating scales. In relevant studies, accuracy is operationalized as the mean honesty rating of truthful messages minus the mean honesty rating of deceptive messages. Because different rating scales are used in different studies, it is necessary to standardize these results before summarizing them. To do so, we divide the mean difference in a study between the rated honesty of truths and lies by a standard deviation from that study. Earlier meta-analyses gave us reason to imagine that rating-scale lie/truth discrimination might be sizeable – yielding mean differences in the range of .86 standard deviations (DePaulo, Zuckerman, & Rosenthal, 1980) or 1.14 standard deviations (Zuckerman, DePaulo, & Rosenthal, 1981).

We found 107 samples of receivers who rated deceptive and truthful messages on multi-point scales. For each of these samples, we computed a standardized difference between means (d). The unweighted mean d was .35 (s = .47). The ds at the first, second, and third quartile were .09, .31, and .67. By fixed-effects methods (Lipsey & Wilson, 2001), the weighted mean d for lie/truth discrimination is .34; 95% confidence interval = .31 to .38. There is statistically significant heterogeneity in the size of these Cohen’s ds, Q(106) = 458.74, p < .01. Receiver sampling error accounts for 21.92% of the observed variance in effect-sizes, and the true standard deviation in these standardized mean differences is .37. It is noteworthy that the level of lie/truth discrimination we find in 107 studies of rated veracity (mean d = .35) is less than half as large as the levels reported in earlier rating reviews (where ds exceeded .85).

Existing summaries led us to suspect that lies might be better discriminated from truths when the discrimination was attempted on multi-point rating scales rather than with dichotomous classifications. To assess this suspicion, we also computed a standardized mean difference for each study in which participants made lie-or-truth classifications. In such cases, the relevant measure is the mean percentage of truthful messages classified as truths minus the mean percentage of deceptive messages classified as truths divided by a standard deviation.

The dichotomous standardized mean differences yielded a weighted mean of .42 in 216 samples from which they could be computed. Values at the first, second, and third quartile were .02 .50 and 1.04. For 61 other samples, no standard deviation in percentage correct lie/truth classifications was reported. There we used the binomial distribution to impute a within-message standard deviation across receivers, and found a weighted mean d of .40.

As these computations indicate, the standardized mean difference in the perceived truthfulness of truths and lies is smaller when receivers use rating scales, rather than when they make lie-or-truth classifications, weighted mean ds = .34 vs. 41; for the difference, Q(1) = 8.86, p < .05. Combining together lie/truth discrimination results from all 384 receiver samples, we find weighted and unweighted mean ds of .39 and .49, respectively. The median d is .39. Standardized mean differences can be converted to Pearson product-moment correlation coefficients. If we convert each d to an r and cumulate the latter in the usual way, we find an unweighted mean accuracy r = .19 and r corresponding to the weighted Fisher’s Z = .21.

Here lie/truth discrimination abilities produce a weighted mean d of approximately .40. This is considerably smaller than the ds of .86, 1.07, and 1.14 reported in earlier rated-veracity reviews. Even so, the ability to discriminate lies from truths at this level should not be dismissed. Many widely cited effects in social psychology are smaller than this one. Indeed, our d of .39 (or r of .21) would rank above the 60^th percentile in size, relative to 474 social psychological effects compiled by Richard, Bond, and Stokes-Zoota (2003).

Percentage Judged True

Deception judgments can have large consequences whether or not they are correct. Thus, it is important to understand factors that may bias the judgments in one direction or another. Vrij (2000) reviewed evidence for a truth bias – receivers’ tendency to err in the direction of judging messages as true.

Researchers reported the percentage of messages receivers classified as true in 207 receiver samples. These are displayed on left side of the stem-and-leaf plot in Figure 1. As Figure 1 shows, the percentage of truth classifications is higher than the percentage of correct classifications, and the percentage of truth classifications is more variable. Percentage truth classifications show an unweighted mean of 56.86% and weighted mean of 55.23%. Each of these values is significantly greater than 50%; for the weighted mean, t(6914) = 46.85, p < .0001. The 95% confidence interval for the weighted mean percentage judged true extends from 54.99% to 55.46% and the true standard deviation across studies in this percentage is 8.13.

Senders succeed in conveying more honesty than dishonesty in these studies. However, the bias thus introduced into receivers’ judgments (of roughly 56% truth judgments) is smaller than the 61% truth judgments reported in a tabulation of 9 studies (Vrij, 2000). Across studies, there is no relationship between the percentage of truth judgments receivers rendered and the percentage of correct lie/truth classifications they achieved, r = -.008.

Stimulus Accuracy

Because an overall accuracy score is computed by averaging the percentage of correct classifications of truthful messages with the percentage of correct classifications of deceptive messages, it may seem informative to analyze separately the two component scores. We will be regarding these two scores as indices of stimulus accuracy for truthful messages and stimulus accuracy for deceptive messages, respectively.

In 207 receiver samples, percentage accuracy rates could be determined for truthful messages and deceptive messages separately. These are the same 207 samples used in our previous tabulation of the truth bias. Unweighted analyses reveal that people correctly classify 61.34% of truthful messages as truthful and 47.55% of deceptive messages as deceptive.

There is variability from study to study in the percentage correct classification of deceptive messages as well as truthful messages (each s =12.51%). The greater the percentage of lies in a study that are correctly classified, the lower is the percentage of truths in that study correctly classified; for the cross-study relationship, r = -.53, p < .0001. Across studies, accuracy at detecting lies shares little variance with accuracy at detecting truths. Any shared accuracy variance is overwhelmed by cross-study differences in suspicion. Thus, cross-study differences result largely from differences in response threshold, rather than differences in discrimination ability.

Response Accuracy

The questions of whether people can identify truths as truths and lies as lies, and of differential rates of success, are important ones. But they tell only part of the story about accuracy at detecting deception. Left unanswered are two parallel questions. Given that a person has judged a statement to be truthful, what is the likelihood that the statement was in fact truthful? And, given that a person has judged a statement to be a lie, what is the likelihood that it was actually a lie? To address these questions, we determined the response accuracy of a receiver’s truth judgments and the receiver’s lie judgments – defining them as the percentage of truth (and of lie) judgments that were correct. Recognizing that response accuracy scores could depend heavily on the baseline percentages of truthful and deceptive messages judged, we restricted our analyses of these measures to receivers who judged an equal number of deceptive and truthful messages. Unweighted means on the relevant 187 samples indicate that judgments of truthfulness are less likely to be accurate than are judgments of deceptiveness, unweighted means = 54.12% vs. 55.84%, t(187)= 5.12, p < .01. There are cross-study differences in the response accuracy of lie and truth judgments (s = 7.75 and 5.84, respectively). Interestingly, the greater the response accuracy of truth judgments in a study, the greater is the response accuracy of lie judgments in that study, and this relationship is strong, r = .80, p < .0001.

Other Accuracy Measures

All of the measures of accuracy that we have considered so far have limitations. Stimulus accuracy measures can be inappropriately affected by variations in judgmental bias, while response accuracy measures can be artifactually affected by variations in deception baserate. In light of these limitations, we analyzed this research literature with several alterative measures of lie/truth discrimination accuracy – including the log odds ratio and d’. These measures have a theoretical advantage over percentage correct, as they are statistically independent of variations in judgmental bias and baserate. We had imagined that these alternative measures might provide distinctive information about people’s average ability to detect lies and give us new insights into cross-study differences in lie detection. They did not.

For one set of analyses, we used methods described in Fleiss (1994) to compute a detection accuracy odds ratio. This was a ratio of the odds that a truthful message was judged to be the truth (rather than a lie) divided by the odds that a deceptive message was judged to be the truth (rather than a lie). Aggregating log odds ratios across the 207 samples in this literature for which requisite data are available, a back-transformation of the mean indicates that the odds of judging a truthful message as the truth is 1.46 times as great as the odds of judging a deceptive message to be the truth (95% CI = 1.41 - 1.51). The back-transformed mean odds of truth detection is 1.65, and the corresponding mean for lie detection is .91. These imply means of 62.30% and 47.53% correct judgments to truthful messages and deceptive messages, respectively – quite close to the stimulus accuracy means of 61.34% and 47.55% directly computable from these samples.

Encouraged by signal detection theorists (e.g., Swets, 1996), we used a binormal method to calculate d’ from each of 207 receiver samples. Here d’ represents a mean difference in apparent honesty between deceptive and truthful messages. The mean d’ in these studies is .24; the median is .22. Although a given d’ can correspond to a number of different percentage correct lie/truth classifications, the maximum percentage correct would occur if the percentage correct judgments of deceptive messages equaled the percentage correct judgments of truthful messages (Walter, 2001). For the mean d’ of .24 in this literature, this maximum is 54.79% -- quite close to the mean of 54.45% correct directly computable from these samples.

Calculations with the odds ratio and d’ corroborate the general conclusion we reached from analyzing percentage correct -- that in the typical research setting lies are discriminated from truths at levels that are slightly better than would be attained by flipping a coin. To determine whether these alternative accuracy measures might give us distinctive information about cross-study differences in lie/truth discrimination, we computed some correlation coefficients across the relevant 207 receiver samples. Results reveal that the three accuracy measures we have been discussing are very highly inter-correlated. As an index of cross-study accuracy differences, percentage correct is virtually interchangeable with the log odds ratio (r = .979) and d’ (r = .988). The latter two measures are barely distinguishable, r = .999. These results should be heartening to the many researchers who have been measuring lie detection accuracy as percentage correct.

Determinants of Accuracy

Thus far, our analysis indicates that individuals have some ability to detect deception. On the average, judges achieve about 54% lie/truth discrimination accuracy. As a percentage, lie/truth discrimination abilities seem poor; but when scaled by cross-judge standard deviations, these abilities appear non-negligible. These are typical results over a variety of receiver samples, sender samples, deception media, types of lies, and contexts. Perhaps under certain conditions judges show high percentage lie/truth discrimination rates; perhaps under other conditions, they show trivial standardized discrimination performances. To assess these possibilities, we now examine various subsets of the research literature on deception judgments. We hope to determine how deception judgments are influenced by six factors: 1) deception medium, 2) motivation, 3) preparation, 4) baseline exposure, 5) interaction, and 6) receiver expertise.

Each of these factors will be assessed in its impact on three different indices of judgment: 1) percentage truth classifications, 2) percentage correct lie/truth classifications, and 3) a standardized difference between the perceived veracity of truths and the perceived veracity of lies. Indices 1) and 2) were coded from studies in which individuals made dichotomous lie/truth classifications and were analyzed with the raw techniques of Bond, Wiitala, and Richard (2003). Index 3), which included results from both lie-or-truth classifications and veracity ratings, was analyzed with standardized fixed effects techniques (Lipsey & Wilson, 2001).

To infer the effects of each factor we consider three forms of evidence: within-study comparisons, between-study comparisons, and statistically adjusted comparisons. We aggregate within-study comparisons for each moderator variable that has been examined within studies. Summaries of relevant experiments provide us with controlled evidence of the impact of moderator variables. Unfortunately, the number of experiments that manipulate a given factor is limited, as is the range of conditions under which it has been examined. Thus, we also assess effects using between-study comparisons. We assess the effect of a person’s motivation to lie, for instance, from a comparison of effect-sizes in studies where participants were motivated to lie with effect-sizes in studies where they were not motivated. Although we can base between-study comparisons on impressive amounts of data, the studies at one level of a moderator variable may differ in any number of ways from the studies at another level. In light of these potential confounds, we also make statistically adjusted comparisons. They gauge the impact of a given moderator variable from a multiple regression analysis that adjusts for the impact of other variables. In particular, our statistically adjusted comparisons of percentage truth classifications and percentage correct lie/truth classifications document the partial effect of a given moderator variable from an inverse variance-weighted multiple regression equation that includes as regressors the six factors enumerated above (deception medium, motivation, preparation, baseline exposure, interaction, and receiver expertise), as well as a control variable indicating whether or not messages were edited prior to presentation. Our statistically adjusted comparisons of ds reflect results from an inverse variance-weighted multiple regression equation that includes the seven regressors just mentioned, as well as an eighth variable that indicates whether deception judgments were rendered as lie-or-truth classifications or on multi-point rating scales.

Let us remind the reader of a framework we bring to deception judgments. In our view, people are harsher in evaluating others’ lies than their own. They stereotype liars as conscience-stricken souls. When asked to judge deception, people consult this stereotype and assess its fit to the person at hand. In general, they are reluctant to label an assertion as deceptive when this judgment would imply that the person offering the assertion was a liar. The inaccurate stereotype and unwanted dispositional implication may help explain why receivers’ judgments are so often inaccurate -- (more specifically) why so many deceptive messages are misclassified as truthful. Our double-standard hypothesis also provides a framework for interpreting the effects of various factors on deception judgments, effects which we now consider.

Deception Medium. Deception can be judged over various media. Some may invite application of a stereotype for inferring deceit, while others encourage reflection. The video medium, we suspect, should encourage use of a liar stereotype. Indeed, if forced to judge deceit from nothing more than a video image, observers have recourse to little other than their stereotypes. Access to verbal content gives judges the option of analyzing issues of veracity in a more thoughtful fashion. Thus, it is of interest to compare detection rates for lies that can only be seen vs. those that can be heard.

Having sketched the relevance of our double-standard framework for interpreting deception attempts in different media, let us mention another theoretical perspective. According to Ekman and Friesen (1969), people should be most successful in their attempts at facial deceit and least successful in lying with the body because they are most motivated and able to control the face.

To assess the impact of deception medium, we identified fifty studies that experimentally manipulated this factor and extracted from these studies 177 pair-wise comparisons of lie/truth discrimination accuracy in one medium vs. another medium. Ninety eight of the comparisons were made on percentage correct lie/truth classifications and 79 on rating scales. Converting each comparison to a standardized mean difference, we conducted a fixed-effects meta-analysis. Results show that lie/truth discrimination accuracy is lower if judgments are made in a video rather than an audiovisual or audio medium (for comparison of video to audiovisual and audio lie/truth discrimination, weighted mean ds = -.44 and -.37, Zs = -15.72 and –9.51 in 58 and 34 experimental comparisons, respectively; each p < .0001). In fact, lie/truth discrimination from video presentations is inferior to discriminations made from written transcripts, weighted mean d = -.28, Z = -4.16, p < .001 in 10 experimental comparisons. The levels of lie/truth discrimination achieved from transcript, audiovisual, and audio presentations do not differ significantly from one another.

We tabulated analogous evidence of receivers’ general tendency to perceive messages as truthful. Results show that messages are perceived as less truthful if judged from a video than an audiovisual or audio presentation, weighted mean d = -.29 and -.34, Zs = -4.26 and –5.79 in 14 and 15 experimental comparisons, respectively; each p < .0001). Messages conveyed in transcripts are judged as less truthful than audiovisual messages and as somewhat more truthful than those presented in video, weighted mean ds = -.32 and +.20, Z = -3.31, p < .01 and Z = 1.94, p = .06, in five and four experimental comparisons, respectively. In perceived truthfulness, audio-based messages do not differ significantly from audiovisual or transcript messages; each p > .10.

To complement these within-study comparisons, we examined medium differences across all of the studies in the research literature. In 195 samples, we have data on percentage truth classifications to messages conveyed in one of three media: video-only, audio-only, or audiovisual. Relevant results appear at the top of Table 2 and suggest that there is a truthfulness bias in judging messages that can be heard. Both audio-only and audiovisual presentations received more than 50% truth judgments. As the within-study comparisons indicated, video-only presentations are less often judged truthful. Medium effects on lie/truth discrimination appear in the bottom two-thirds of the table. Corroborating the within-study evidence, these comparisons show that discrimination is poorer for video-only messages than for messages presented in an audio-only or audiovisual medium.

From our double-standard framework, we interpret these results as follows – that the usual stereotype of a liar is largely visual, hence is most strongly evoked by video images of people speaking. Those who can be viewed as tormented are judged to be lying; but apparent torment reflects many factors other than deceit.

Ekman and Friesen (1969) hypothesized that there are more deception cues in the body than the face. To examine this possibility, we divided the video-based deception attempts into ones that provided the receiver with cues from only the face (k = 15), only the body (k =9), or the face plus the body (k =29). Results provide only partial support for the Ekman and Friesen (1969) formulation. Consistent with that formulation, attempts at lie detection are unsuccessful when receivers see only the sender’s face; however, detection efforts are similarly unsuccessful when receivers see only the liar’s body (weighted mean accuracy d = .01, -.15, and +.12 for face, body, and both, respectively).

Motivation. Deception studies are criticized when research participants have no incentive to be believed. Critics note that a lack of motivation may influence participants’ believability. To address this issue, we divided the research literature into studies in which participants had little (or no) motivation to be believed and those in which they had higher motivation.

DePaulo and her colleagues (e.g., DePaulo, Stone, & Lassiter, 1985) have hypothesized that senders are undermined by their efforts to get away with lying. In DePaulo’s motivational impairment hypothesis, the truths and lies of highly motivated senders will be more easily discriminated than those of unmotivated senders unless receivers have access to nothing but a transcript of the sender’s remarks.

For a controlled assessment of this hypothesis, we identified 20 studies that experimentally manipulated sender motivation, extracted from those studies 42 distinguishable motivation effects on lie/truth discrimination, and measured each effect as a standardized mean difference. Consistent with the motivational impairment hypothesis, experimental evidence shows that lies are easier to discriminate from truths if they are told by motivated rather than unmotivated senders (for impact of motivation, weighted mean d = .171, Z = 7.10, p < .0001).

The double-standard hypothesis has a different implication for understanding the impact of motivation on deception judgments. People who are afraid of being disbelieved may come to resemble the stereotypic liar. If so, they are likely to be judged deceptive. From this perspective, it should matter little whether or not a highly motivated speaker is lying. What matters is the speaker’s fear of being disbelieved. High motivation would rarely make a person feel guilty or ashamed for lying; indeed, high stakes should make it easy to rationalize deceit.

For between-study evidence relevant to this perspective, see the top third of Table 3. Consistent with the double-standard hypothesis, motivation to be believed reduces a speaker’s apparent honesty. Perhaps motivation makes people resemble a visible stereotype of the liar. If so, motivational effects on credibility might be most apparent on video-based judgments. To assess this possibility, we examined the impact of motivation on lie- and truth-tellers’ believability in video, audio, and audiovisual media. Between-study comparisons reveal that motivation significantly reduces senders’ video and audiovisual appearance of truthfulness. For example, unmotivated and motivated senders are classified as truthful by 54.44% and 46.84% of receivers who see them in video-only presentations; t’(95) = 7.17, p < .001. However, motivation has no effect on how truthful a sender sounds; t’(137) = 1.31, n.s.

The bottom two-thirds of Table 3 displays between-study evidence on sender motivation and lie/truth discrimination. Here it does not appear that motivation makes liars easier to detect.

Preparation. Sometimes the need to lie appears without warning, and people are unprepared for the deceptions they attempt. On other occasions, the need has been anticipated, and a line has been prepared. In principle, the opportunity to prepare might influence a liar’s success.

To examine this possibility, we identified fifteen studies that experimentally manipulated a sender’s time to prepare lies. These studies reported 24 experimental effects of sender preparation on the accuracy of lie/truth judgments and 10 experimental effects on the sender’s general tendency to appear truthful. A fixed-effects standardized meta-analysis shows that receivers achieve higher lie/truth detection accuracy when judging unplanned, rather than planned messages (weighted mean d = -.144, Z = 4.49, p < .01), and that planned messages appear more truthful than messages which were unplanned (weighted mean d = .133, Z = 2.35, p < .05).

Relevant between-study evidence is displayed in Table 4. Although the results there for judgment accuracy are mixed, they suggest that it may be harder to discriminate deceptive from truthful messages when the messages are planned. Unlike the within-study evidence, between-study comparisons suggest that planned messages appear slightly less honest than spontaneous messages.

Baseline exposure to sender. The current meta-analysis focuses on judgments of deception among strangers. Even so, we included in the analysis 38 samples in which perceivers were exposed to a target before making judgments of that target. We also included 28 samples in which perceivers judged a given target eight or more times and four samples in which perceivers made a forced choice between a target’s lie and that same target’s truth. For purposes of the analyses below, all of these receivers were deemed to have received a baseline exposure to the target.

For a controlled analysis, we identified 21 experimental comparisons of the detection of a target’s messages by judges who had (vs. judges who had not) been previous exposed to that target. All of these comparisons were made on percentage correct lie/truth judgments. Results indicate that baseline exposure improves lie/truth discrimination: receivers achieve a mean of 55.91% accuracy when given a baseline exposure vs. 52.26% accuracy in the absence of any exposure, t’(364) = 6.37, p < .01.

Between-study evidence on the impact of baseline exposure is displayed in Table 5. Results there suggest that baseline exposure may improve judgmental accuracy. At the same time, senders who are familiar to the receiver are likely to be given the benefit of the doubt, as results on the percentage of truth judgments indicates. Consistent with our double-standard framework, people are reluctant to imply that someone familiar to them is a liar.

Interaction. In many studies, people lie while alone or in the presence of a passive experimenter. In other studies, people are involved in social interactions when lying. Sometimes, the interaction partner is attempting to judge the liar’s veracity; and on other occasions, a third party may be making this judgment. The latter occurs, for example, when the interaction partner is the experimenter and the third party is the receiver making judgments from a videotape. In principle, interaction might influence one’s success at lying. Interaction might, for example, impose cognitive demands on the liar (Buller & Burgoon, 1996).

We found eleven studies that experimentally manipulated whether senders were interacting with the receiver or with a third party. Results indicate no significant difference in lie/truth discrimination by interaction partners (vs. third-party observers), weighted mean ds = .286 vs. .209, Z = 1.41, n.s. We also tabulated evidence within five studies of receivers’ general tendency to perceive senders as truthful. Results show that individuals are judged to be more truthful by their interaction partners than by third-party observers; for this comparison, weighted mean d = .26, Z = 4.10, p < .0001.

For between-study evidence on the impact of interaction, see Table 6. There it is again clear that receivers are inclined to judge their interaction partners as truthful. Overall patterns in the literature suggest that third-party observers are better than interaction partners at discriminating lies from truths. In our view, the reluctance to attribute deception to interaction partners results from an unwanted dispositional implication – of insinuating that the partner is a liar.

Receiver expertise. In most research, college students function as the judges of deception. Perhaps people who had more experience would be better at judging deceit. To assess this possibility, we identified studies of deception experts. These are individuals whose occupations expose them to lies. They include law enforcement personnel, judges, psychiatrists, job interviewers, and auditors – anyone whom deception researchers regard as experts.

In 19 studies, expert and non-expert receivers judged the veracity of the same set of messages. From these studies, we extracted 20 independent expert/non-expert comparisons, and expressed each as a standardized mean difference. This cumulation yields no evidence that experts are superior to non-experts in discriminating lies from truths; weighted mean d = -.025, 95% confidence interval = -.105 to +.055. Indeed, the direction of the within-study difference favors higher non-expert accuracy, though this difference is not statistically significant, Z = -.61, n.s. Within-study comparisons also reveal no statistically significant difference between experts and non-experts in the tendency to perceive others as truthful; weighted mean percentage truth judgments = 54.09% and 55.74% for experts and non-experts, respectively; t’(246)=1.41.

For a broader assessment of experts’ deception judgments, see Table 7. From the between-study evidence, it would appear that experts are more skeptical than non-experts, being less inclined to believe that people are truthful. Having been targets of deceit in their professional roles, experts may have surmounted the usual reluctance to imply that people are liars. If raw between-study comparisons suggest that experts may be better than non-experts at discriminating lies from truths, it is clear that experts are not good lie detectors. On the average, they achieve less than 55% lie/truth discrimination accuracy. In any case, experts’ apparent superiority in lie/truth discrimination disappears when means are statistically adjusted.

Publication Status. Lie detection results might influence the likelihood of a research project being published. To assess this possibility, we did a few other analyses. These reveal no statistically significant differences between published and unpublished studies in lie/truth discrimination performances. For example, the weighted mean percentage correct lie/truth classifications is 53.19% in published studies and 53.75% in unpublished studies, t’(872)=1.49, n.s. Truthfulness biases were, however, stronger in unpublished research; weighted mean percentage truth classifications = 56.75% vs. 54.27% in unpublished vs. published research, t’(498)=4.75, p < .001.

Discussion

Having captivated human imagination for millenia, deception was destined to attract psychological investigators. Our goal has been to synthesize their research – more specifically, to quantify people’s ability to detect deceit from behavior. Here we summarize the findings of our meta-analysis, discuss the literature in light of a double-standard framework, and note limitations in the existing evidence.

Meta-analytic Findings

How successful are people at duping others? How often do people detect others’ deception attempts? To address these questions, psychologists arrange for people to make truthful and deceptive statements and for others to classify these statements as truths or lies. Across hundreds of experiments, typical rates of lie/truth discrimination are slightly above 50%. For the grand mean, 54% is a reasonable estimate.

Having noted that the average person discriminates lies from truths at a level slightly better than s/he could achieve by flipping a coin, let us also note this ability corresponds to a nontrivial standardized effect size. In producing a mean difference of approximately .40 standard deviations in judgments of lies vs. truths, typical detection abilities are larger than 60% of the research phenomena studied by social psychologists (Richard, Bond, & Stokes-Zoota, 2003).

Our finding of a 54% lie/truth discrimination rate represents an average of correct judgments to deceptive messages and truthful messages. It is clear that truthful messages are more often judged correctly than deceptive messages; hence the percentage of correct judgments to messages encountered in any real-world setting may depend on the base rate of deception there. In a setting where virtually no lies were told, the research literature would suggest a detection rate of roughly 60%; while in a situation where virtually every statement was a lie, a detection rate of, say, 48% might be expected (cf. Levine, Park, & McCornack, 1999). These estimates assume that there is no cross-situational correlation between observers’ tendency to infer deception in a setting and the actual rate of lying there. More likely, deception base rates enter into a tactical calculus. As observers have intuitions about the frequency of deception in different situations, liars have intuitions too. If the latter can choose where to attempt their deceptions, they should opt for settings in which targets are most trusting.

Like earlier reviewers, we find that people are more inclined to judge deceptive messages as truthful than truthful messages as deceptive. No doubt, receivers contribute to this truth bias, but senders’ contributions should also be acknowledged. When people try to appear truthful, their efforts are rewarded – the accumulated literature shows. The relative impact of senders and receivers on the truth bias remains to be determined. In the meantime, the present contribution is to document the magnitude of this effect. Across 206 studies, people render a mean of some 56% truth judgments. However, this figure may understate the presumption of truth telling in real life. If in their daily interactions people accept without reflection much of what they hear, in the laboratory they are forced to make veracity judgments. Thus, researchers circumvent some of the usual impediments to inferring deceit – social norms that discourage skepticism, liars’ tactics for pre-empting suspicion, and a cognitive inertia that would be disrupted by critical inquiry (Levine, Park, & McCornack, 1999).

We see a pattern in this research literature. In their reading of the literature, scholars find an unwanted implication -- that people can barely discriminate lies from truths. Heirs to the moralistic tradition, scholars resist this implication by identifying a feature of researchers’ methods that could in principle explain low lie/truth discrimination rates. They label the feature an artifact, correct the error, run a study, and announce that their findings are uniquely valid. Sometimes, the methodological correction yields a higher than average detection rate, and sometimes it does not. Never, however, has this quest for accuracy yielded levels of lie detection that would be of much practical use. Occasionally, a researcher finds a detection rate of 70% (or so) and proclaims a momentous discovery. However, those rates occur on tests that include only a small number of messages and are attained by only a subset of the receivers (or on a subset of the tests) studied. From a meta-analytic perspective, random variation is the most plausible explanation for the occasionally high detection rate, as the funnel pattern in Figure 2 suggests.

Rather than marveling at the outliers in this literature, we are more impressed by the regularity of the results obtained. Despite decades of research effort to maximize the accuracy of deception judgments, detection rates rarely budge. Professionals’ judgments, interactants’ judgments, judgments of high-stakes lies, judgments of unsanctioned lies, judgments made by long-term acquaintances – all reveal detection rates within a few points of 50%. We wonder if it is premature to abort the quest for 90% lie detection and accept the conclusion implied by the first 384 research samples – that to people who must judge deception in real time with no special aids, many lies are undetectable.

Although rates of lie detection vary within a narrow range, the variation is not random. Some factors facilitate lie/truth discrimination, and others impede it – our meta-analytic results confirm. The medium in which deception is attempted affects its likelihood of detection – lies being more detectable when they can be heard. By contrast, facial behaviors provide no indication of a speaker’s veracity, corroborating the theory that the face is well controlled (Ekman & Friesen, 1969). Ekman and Friesen also suggested that bodily behaviors go uncontrolled, hence should be indicative of deceit. Unfortunately, the latter hypothesis has so rarely been tested that its validity remains unknown.

A more recent perspective (Buller & Burgoon, 1996) emphasizes the role of social interaction in deception judgments. The accumulated research suggests that lies told in the midst of social interaction are spotted by on-lookers, while they are fooling the liar’s interaction partner. However, controlled experiments show no difference in lie detection by interaction partners, as opposed to on-lookers. As common sense might have predicted, judges achieve better lie/truth discrimination if they have a baseline exposure to the sender and if the sender is unprepared. The accumulated evidence suggests that people who are motivated to be believed look deceptive, whether or not they are lying. Expert judges may be slightly more skeptical than novices. Relative to novices, experts may (or may not) be better at lie/truth discrimination; in any case, they make many mistakes.

The Double Standard

Having reviewed the research literature on deception judgments and cataloged some factors that influence detection accuracy, let us note the relevance of our favored framework for understanding this subject – our assumption that people judge others’ deceptions more harshly than their own.

We do not regard the current meta-analysis as a test of the notion of a double standard. In our view, no test for so obvious an idea is needed – though relevant evidence can be found in primary research (Gordon & Miller, 2000). Instead, we begin with the premise that people construe others’ lies more critically than their own and explore the implications of this premise for understanding research findings.

Indignant at the prospect of being duped, people project onto the deceptive a host of morally fueled emotions – anxiety, shame, and guilt. Drawing on this stereotype to assess others’ veracity, people find that the stereotype seldom fits. In underestimating the liar’s capacity for self-rationalization, judges’ moralistic stereotype has the unintended effect of enabling successful deceit. Because deceptive torment resides primarily in the judge’s imagination, many lies are mistaken for truths. When torment is perceived, it is often not a consequence of deception but of a speaker’s motivation to be believed. High stakes rarely make people feel guilty about lying; more often, they allow deceit to be easily rationalized. When motivation has an impact, it is on the speaker’s fear of being disbelieved; and it matters little whether or not the highly motivated are lying. The impact of motivation is most evident when judges can see the speaker’s resemblance to a visual stereotype of the liar.

People are critical of lies, unless the lies are their own. To maintain an exception for themselves, judges may sometimes need to excuse lying by others. As the research literature shows, people avoid attributing deception to others with whom they are familiar – whether from a live interaction, or a long-term relationship (Anderson, Ansfield, & DePaulo, 1999). Judges may also be loath to perceive as liars people who resemble the judge. Perhaps the truth bias we observe in this literature represents an extension of the self bias to others who are reminiscent of the self. In this view, the bias reflects the similarity of the deceivers in this research literature to their judges – often, the two are students at the same University. Maybe there would be less bias in judgments made of dissimilar others. As we have noted, deception researchers find that expert judges are willing to imply that others are liars. What we have not noted is a procedural detail -- that these experts are rarely sitting in judgment of their peers; instead, they are judging members of other groups. Self biases do not extend to outsiders.

The judges in this research literature are given the goal of achieving 100% accuracy, and their failure to attain this objective has been widely lamented. The senders in this research literature are also given a goal: to convey an impression of honesty 100% of the time. Results show that research participants disbelieve nearly 50% of senders’ deception attempts and nearly 40% of their attempts at truth telling. Although in the rough actuarial aggregate of deception research liars fail as often as detectors, deception failures have rarely been discussed. Let us comment on these failures from a double standard perspective.

Liars who are often judged deceptive should come to learn that their stratagems have been penetrated. Thus, it may seem paradoxical that the average person lies several times a day (DePaulo, Kashy, Kirkendohl, Wyer, & Epstein, 1996). Evidently, most lies are little, and the consequences of detection benign. In the interest of interacting smoothly, the liar and judge conspire to preserve a fiction (DePaulo et al, 2003).

A few lies involve high stakes: large benefits to the liar and large costs to the dupe. Moralists focus on these big bad lies. The research literature has explored judgments made at the time deception is attempted, judgments that could pre-empt the payoffs liars pursue. However, research reveals that many people avoid being caught in the act of lying; hence, it is important to explore the likely course of subsequent events.

High-stakes deceptions are motivated by non-correspondent outcomes, one person seeking advantage at another’s expense. There are costs of being duped, and these should impose limits on the dupe’s naiveté. Some lies are discovered well after they have been told (Park, Levine, McCornack, Morrison, & Ferrara, 2002). Then the dupes become indignant. They retaliate by shunning their exploiter and publicizing the liar’s duplicity. As a consequence, people who are most successful in the short-term perpetration of lies have trouble maintaining relationships. Moralists have opined that skilled liars are worse relationship partners than highly honest folk. Let us suggest that skilled liars may also be worse partners than people whose lies are transparent (Andrews, 2002). Inept liars pose no threat to their partners insofar as their deception attempts fail before any damage is done. This line of reasoning suggests that skill at high-stakes deception may be an interpersonal liability and that so-called deception failures are in the long run adaptive.

Maybe the craftiest can benefit from lying. Cognizant of the dispositional nature of moral attributions (Reeder, 1993), they cultivate reputations for honesty by telling the truth on trivial matters, while noting advantages that fibbing might have conferred. Then when deceit promises the largest reward, others will have been lulled into an unwarranted trust (Sternglanz, 2003). Having laid the tactical groundwork, liars must nonetheless recognize that deceptions may ultimately be exposed. In the moment of lying, the shrewdest affect a distancing from their falsehoods, so that they can later disavow the lies. For deception to show long-term profitability, reputational damage must be contained.

Limitations in the Evidence

Commentators have criticized research on deception judgments, pointing to ways in which the lies studied in the research literature differ from the lies of most interest to the critic. Those who are interested in high-stakes lies (Ekman, 2001) note that many experimental deceptions are trivial. Those who are interested in deceptive interactions (Burgoon & Buller, 1996) denounce experimentally constrained lies. Legal scholars (e.g., Fisher, 1997) note aspects of the forensic world that are not reproduced in research contexts.

Deception researchers have tried to accommodate critics’ reservations. They have studied murderers’ lies and lies that could harm children (Lusby, 1999; Vrij & Mann, 2001a), lies to lovers and deceit during criminal interrogations (Anderson, Ansfield, & DePaulo, 1999; Davis, Markus, Walters, Vorus, & Conners, In press). Researchers have studied naturalistic deceptive interactions and jurors’ credibility judgments (Frank, 1989; Park, Levine, Harms, & Ferrara, 2002). In light of these efforts, we find no merit in blanket dismissals of this research literature as trivial, asocial, and irrelevant.

We ourselves have reservations about the literature on deception judgments, concerns that have not (we think) been addressed. To illuminate lie detection from language and behavior, psychologists have excluded from their research other potential cues to deception. They have restricted the time span over which issues of deception can be pondered, blinded judges to the motivational contingencies surrounding deceit, and neutralized naturally occurring correlates of the propensity to lie.

In experiments, judges encounter a message and must judge the veracity of that message on the spot with no time to gather additional information. Outside the laboratory, additional information is important. When asked to describe their discovery of a lie, people rarely state that the discovery was prompted by behaviors displayed at the time of the attempted deception. Rather, they say that lie detection took days, weeks, or even months, and involved physical evidence or third parties (Park, Levine, McCornack, Morrison, & Ferrara, 2002). Surely, motivational information conditions real-world deception judgments – when, for instance, jurors discount expert testimony after learning that the expert received a fee (Hilton, Fein, & Miller, 1993). In venues of frequent deception, people may base their veracity judgments more strongly on perceived incentives than any behavioral information. People differ widely in the propensity to lie (Kashy & DePaulo, 1996), and this individual difference may be discernable (Bond, Berry, & Omar, 1994). Researchers bypass naturally occurring correlates of deceptiveness by compelling lies from every experimental participant – even those who are loath to lie. Future studies will be needed to explore the impact on lie detection of these and other forms of extra-behavioral information. Perhaps the 90% lie detection barrier will someday be broken.

In the meantime, we have accumulated knowledge about judgments of deception from speech content and behavior. Yes, people often fail in their efforts to divine deception, and this raises questions about the American legal system, where jurors are responsible for detecting lies. It is important also to note that research participants often fail when trying to dupe others. Perhaps it would be unsurprising if liars and would-be detectors had arrived at an equilibrium. If liars were much better, truth telling would be less common; if detectors were much better, few lies would be attempted.

References

Anderson, D. E., Ansfield, M.E., & DePaulo, B. M. (1999). Love’s best habit: Deception in the context of relationships. In Philippot, P., & Feldman, R.S. (Eds.) Social context of nonverbal behavior (pp. 372-409). New York: Cambridge University Press.

Anderson, N.H. (1968). Likeableness ratings of 555 personality-trait words. Journal of Personality and Social Psychology, 9, 272-279.

Andrews, P.W. (2002). The influence of postreliance detection on the deceptive efficacy of dishonest signals of intent: Understanding facial clues to deceit as the outcome of signaling tradeoffs. Evolution and Human Behavior, 23, 103-121.

Bok, S. (1989). Lying: Moral choice in public and private life. New York: Vintage Books.

Bond, C. F., Jr., Berry, D. S., and Omar, A. (1994). The kernel of truth in judgments of deceptiveness. Basic and Applied Social Psychology, 15, 523-534.

Bond, C.F., Jr., & Robinson, M. (1988). The evolution of deception. Journal of Nonverbal Behavior, 12, 295-307.

Bond, C.F., Jr., Wiitala, W., & Richard, F.D. (2003). Meta-analysis of raw mean differences. Psychological Methods, 8, 406-418.

Buller, D.B., & Burgoon, J.K. (1996). Interpersonal deception theory. Communication theory, 6, 203-242.

Davis, M., Markus, K.A., Walters, S.B., Vorus, N., & Connors, B. (In press). Behavioral cues to deception vs. topic incriminating potential in criminal confessions. Law and Human Behavior.

DePaulo, B.M., Ansfield, M.E., & Bell, K.L. (1996). Theories about deception and paradigms for studying it: A critical appraisal of Buller and Burgoon’s interpersonal deception theory and research. Communication Theory, 3, 297-310.

DePaulo, B.M., Kashy, D.A., Kirkendol, S.E., Wyer, M.M., & Epstein, J.A. (1996). Lying in everyday life. Journal of Personality and Social Psychology, 70, 979-995.

DePaulo, B.M., Lindsay, J.J., Malone, B.E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin, 129, 74-118.

DePaulo, B.M., Stone, J.L., & Lassiter, G.D. (1985). Deceiving and detecting deceit. In B.R. Schenkler (Ed.), The self and social life (pp. 323-370). New York: McGraw-Hill.

DePaulo, B.M., Zuckerman, M., & Rosenthal, R. (1980). Humans as lie detectors. Journal of Communication, 30, 129-139.

Ekman, P. (2001). Telling lies: Clues to deceit in the marketplace, politics, and marriage (Third edition). New York: Norton.

Ekman, P., & Friesen, W.V. (1969). Nonverbal leakage and clues to deception. Psychiatry, 32, 88-105.

Fisher, G. (1997). The rise of the jury as lie detector. Yale Law Journal, 103, 575-713.

Fleiss, J.L. (1994). Measures of effect size for categorical data. In H. Cooper and L. Hedges (Eds) The handbook of research synthesis (pp. 245-260). New York: Russell Sage.

Global Deception Research Team (In press). A world of lies. Journal of Cross Cultural Psychology.

Gordon, A.K., & Miller, A.G. (2000). Perspective differences in the construal of lies: Is deception in the eye of the beholder? Personality and Social Psychology Bulletin, 26, 46-55.

Hedges, L.V., & Vevea, J.L. (1998). Fixed- and random-effects models in meta-analysis. Psychological Methods, 3, 486-504.

Hilton, J.L., Fein, S., & Miller, D.T. (1993). Suspicion and dispositional inference. Personality and Social Psychology Bulletin, 19, 501-512.

Hubbell, A.P., Mitchell, M.M., & Gee, J.C. (2001). The relative effects of timing of suspicion and outcome involvement on biased message processing. Communication Monographs, 68, 115-132.

Judicial Committee on Modern Jury Instructions for the Eighth Circuit (2002). Manual of modern criminal jury instructions. Published on-line at www.are.uscourts.gov/jury/crjury2002.pdf.

Kalbfleisch, P.J. (1990). Listening for deception: The effects of medium on accuracy of detection. In R.N. Bostrom (Ed.) Listening behavior: Measurement and application (pp. 155-176). New York: Guilford Press.

Kashy, D.A., & DePaulo, B.M. (1996). Who lies? Journal of Personality and Social Psychology, 70, 1037-1051.

Kraut, R. (1980). Humans as lie detectors: Some second thoughts. Journal of Communication, 30, 209-216.

Levine, T.R., Park, H.S., & McCornack, S.A. (1999). Accuracy in detecting truths and lies: Documenting the "veracity effect". Communication Monographs, 66, 125-144.

Light, R.J., Singer, J.D., & Willett, J.B. (1994). The visual presentation and interpretation of meta-analyses. In H. Cooper and L.V. Hedges (Eds.) The handbook of research synthesis (pp. 439-453). New York: Russell Sage.

Lusby, David J. (1999). Effects of motivation to lie and sanctioned deception of the accuracy of observers’ veracity judgments. Unpublished M.A. thesis, SUNY-Buffalo.

Mattson, M., Ryan, D.J., Allen, M., & Miller, V. (2000). Considering organizations as a unique interpersonal context for deception detection: A meta-analytic review. Communication Research Reports, 17, 148-160.

Miller, G.R., & Stiff, J.B. (1993). Deceptive communication. Newbury Park, CA: Sage.

O’Sullivan, M. (2003). The fundamental attribution error in detecting deception: The boy-who-cried-wolf effect. Personality and Social Psychology Bulletin, 29, 1316-1327.

Park, H.S., & Levine, T.R. (2001). A probability model of accuracy in deception detection experiments. Communication Monographs, 68, 201-210.

Park, H.S., Levine, T.R., McCornack, S.A., Morrison, K., & Ferrara, S. (2002). How people really detect lies. Communication Monographs, 69, 144-157.

Reeder, G.D. (1993). Trait-behavior relations and dispositional inference. Personality and Social Psychology Bulletin, 19, 586-593.

Richard, F.D., Bond, C.F., Jr., & Stokes-Zoota, J.J. (2003). One hundred years of social psychology quantitatively described. Review of General Psychology, 7, 331-363.

Saxe, L. (1991). Lying: Thoughts of an applied social psychologist. American Psychologist, 46, 409-415.

Sternglanz, R. W. (2003). Exoneration of serious wrongdoing via confession to a lesser defense. Unpublished Ph..D. dissertation, University of Virginia, Charlottesville, VA.

Swets, J.A. (1996). Signal detection theory and ROC analysis in psychology and diagnostics. Mahwah, NJ: Erlbaum.

Trovillo, P.V. (1939). A history of lie detection. Journal of Criminal Law and Criminology, 29, 848-881.

Vrij, A. (2000). Detecting lies and deceit: The psychology of lying and the implications for professional practice. New York: John Wiley.

Vrij, A., Edward, K., Roberts, K., & Bull, R. (2000). Detecting deceit via analysis of verbal and nonverbal behavior. Journal of Nonverbal Behavior, 24, 239-263.

Vrij, A., & Mann, S. (2001a). Telling and detecting lies in a high-stake situation: The case of a convicted murderer. Applied Cognitive Psychology, 15, 187-203.

Vrij, A., & Mann, S. (2001b). Who killed my relative? Police officers' ability to detect real-life high-stake lies. Psychology, Crime, & Law, 7, 119-132.

Walter, S.D. (2002). Properties of the summary receiver operating characteristic (SROC) curve for diagnostic test data. Statistics in Medicine, 21, 1237-1256.

Zuckerman, M., DePaulo, B.M., & Rosenthal, R. (1981). Verbal and nonverbal communication of deception. In L. Berkowitz (Ed.) Advances in experimental social psychology (Volume 14, pp. 1-60). New York: Academic Press.

Table 1

Characteristics of the Research Literature

Quantitative Variables

Variable Minimum Maximum Mean Median s

Number of Senders 1 200 22.45 16.00 22.63

Number of Receivers 1 816 63.65 41.50 70.56

Messages per Receiver 1 416 31.89 16.00 44.50

Message Duration (sec.) 2 1200 110.63 52.00 173.16

Categorical Variables

Variable # (%) of Receiver Samples Percent Coding Agreement

Deception Medium 91.3%

Video 47 (12.2%)

Audio 42 (10.9%)

Audiovisual 262 (67.4%)

Other 22 (4.9%)

Within-receiver manipulation 11 (4.4%)

Sender motivation 89.5%

No motivation 214 (55.7%)

Motivation 153 (39.8%)

Within-receiver manipulation 17 (4.4%)

Sender preparation time 81.1%

None 196 (51.0%)

Some 165 (43.0%)

Within-receiver manipulation 23 (6.0%)

Baseline exposure 91.3%

No exposure 360 (93.7%)

Exposure 20 (5.2%)

Within-receiver manipulation 4 (1.1%)

Sender interaction 100%

None 127 (33.1%)

Interaction with receiver 33 (8.6%)

Interaction with another 224 (58.3%)

Receiver expertise 100%

Not expert 338 (88.0%)

Expert 46 (12.0%)

Figure 1

Stem and leaf plots

Mean percentage truth judgments Mean percentage correct lie/truth judgments

(k = 207) (k = 292)

Leaves Stem Leaves

0 9

666 8

322 8

110 8

98 7

76 7

5544 7

2 7 3

110000 7 11

98888888 6

7777 6 667777

55555554 6 444455

333222 6 22222222333

1111000000000 6 0000000000000111111

9999999988888888 5 88888888888888888999999999999999

7777777777777666666 5 6666666677777777777777777

555555544444444444 5 4444444444444444444444445555555555555555555555

3333333333333332222222222222 5 22222222222222233333333333333333333

1111100000000 5 000000000000000000000111111111111111111111

99999988888 4 888888888889999999999999

777766666666666666666 4 66666667777777777777

5544 4 4455555

33222 4 2222

4 0011

9 3 99

3 7

55 3 5

10 3 1

9 2

77 2

Figure 2

Mean Percent Correct by Number of Judgments

0

10000

Appendix A

Studies Included in the Meta-analysis

1. Al-Simadi, F.A. (2000). Detection of deceptive behavior: A cross-cultural test. Social Behavior & Personality, 28, 455-461.

2. Anderson, D.E. (1999). Cognitive and motivational processes underlying truth bias. Unpublished Ph.D. dissertation, University of Virginia.

3. Anderson, D.E., DePaulo, B.M., & Ansfield, M.E. (2002). The development of deception detection skill: A longitudinal study of same-sex friends. Personality and Social Psychology Bulletin, 28, 536-543.

4. Ask, K., & Granhag, P.A. (2003). Individual determinants of deception detection performance: need for closure, attributional complexity, and absorption. Goteborg Psychological Reports, 33, 1-13.

5. Atmiyanandana, V. (1976). An experimental study of the detection of deception in cross-cultural communication. Unpublished doctoral dissertation, Florida State University.

6. Bailey, J.T. (2002). Detecting deception when motivated: The effects of accountability and training on veracity judgments. Unpublished M.S. thesis, Ohio University.

7. Bauchner, J.E., Kaplan, E.A., & Miller, G.R. (1980). Detecting deception: The relationship of available information to judgmental accuracy in initial encounters. Human Communication Research, 6, 253-264.

8. Berger, R.E. (1977). Machiavellianism and detecting deception in facial nonverbal communication. Towson State University Journal of Psychology, 1, 25-31.

9. Billings, F.J. (2004). Psychopathy and the ability to deceive. Unpublished Ph.D. dissertation, University of Texas at El Paso.

10. Blair, J.P., & McCamey, W.P. (2002). Detection of deception: An analysis of the behavioral analysis interview technique. Illinois Law Enforcement Executive Forum, 2, 165-169.

11. Bond, C.F., Jr., & Atoum, A.O. (2000). International deception. Personality and Social Psychology Bulletin, 26, 385-395.

12. Bond, C.F., Jr., & Fahey, W.E. (1987). False suspicion and the misperception of deceit. British Journal of Social Psychology, 26, 41-46.

13. Bond, C.F., Jr., Kahler, K.N., & Paolicelli, L.M. (1985). The miscommunication of deception: An adaptive perspective. Journal of Experimental Social Psychology, 21, 331-345.

14. Bond, C.F. Jr., Lashley, B.R., & Kirk, C.T. (2002). Responding to deception. Unpublished manuscript, Texas Christian University.

15. Bond, C.F., Jr., Omar, A., Mahmoud, A., & Bonser, R.N. (1990). Lie detection across cultures. Journal of Nonverbal Behavior, 14, 189-204.

16. Bond, C.F., Jr., Omar, A., Pitre, U., Lashley, B.R., Skaggs, L.M., & Kirk, C.T. (1992). Fishy-looking liars: Deception judgment from expectancy violation. Journal of Personality and Social Psychology, 63, 969-977.

17. Bond, C.F., Jr., Paulson, R.M., & Thomas, B.J. (2003). Multiple-audience messages. Unpublished data, Texas Christian University.

18. Bond, C.F., Jr., & Prestwood, D. (2004). Investigating deceptive claims. Unpublished manuscript, Texas Christian University.

19. Bond, C.F., Jr., Thomas, B.J., & Paulsen, R.M. (2004). Maintaining lies: The multiple-audience problem. Journal of Experimental Social Psychology, 40, 29-40.

20. Bond, G.D., Malloy, D.M., Arias, E.A., Nunn, S.N., & Thompson, L.A. (2005). Lie-biased decision making in prison. Communication Reports, 18, 1-11.

21. Bond, G.D., Malloy, D.M., Thompson, L.A., Arias, E.A., & Nunn, S.N. (2004). Post-probe decision making in a prison context. Communication Monographs, 71, 269-283.

22. Bond, G.D., Thompson, L.A., & Malloy, D.M. (2005). Vulnerability of older adults to deception. Psychology and Aging, 20, 60-70.

23. Boone, R.T., Blumenthal, J.A., Simon, T., Cunningham, J.G., & Tucker, J.S. (1998). Discriminating the truth: The effect of presentation, order, and gender on the detection of deception in an adversarial dispute. Unpublished manuscript, Brandeis University.

24. Brand, R.J, Hodges, S.D, & Williams, J.L. (2003). Sex differences in encoding sexual attraction. Paper presented at the annual meeting of the American Psychological Society.

25. Brandt, D.R., Miller, G.R., & Hocking, J.E. (1980a). Effects of self-monitoring and familiarity on deception detection. Communication Quarterly, 28, 3-10.

26. Brandt, D.R., Miller, G.R., & Hocking, J.E. (1980b). The truth-deception attribution: Effects of familiarity on the ability of observers to detect deception. Human Communication Research, 6, 99-110.

27. Brandt, D.R., Miller, G.R., & Hocking, J.E. (1982). Familiarity and lie detection: A replication and extension. Western Journal of Speech Communication, 46, 276-290.

28. Briggs, J.R. (1992). Counselor assessments of honest and deceptive clients (honest clients). Unpublished Ph.D. dissertation, Ball State University.

29. Buchanan, J.N. (1998). UJOS as lie detectors: Exploring University Judicial Officers’ judgments of deception. Unpublished Ph.D. dissertation, Florida State University.

30. Buller, D.B., & Hunsaker, F.G. (1995). Interpersonal deception: XIII. Suspicion and the truth-bias of conversational participants. In J. Aitken (Ed.) Interpersonal communication processes reader (pp. 237-257). Westland, MI: McNeil.

31. Buller, D.B., Strzyzewski, K.D., & Hunsaker, F.G. (1991). Interpersonal deception: II. The inferiority of conversational participants as deception detectors. Communication Monographs, 58, 25-40.

32. Burgoon, J.K., Blair, J.P., & Strom, R.E. (2005). Heuristics and modalities in determining truth versus deception. Paper presented at the 38^th Hawaii International conference on system sciences.

33. Burgoon, J.K., Buller, D.B., Dillman, L., & Walther, J.B. (1995). Interpersonal deception: IV. Effects of suspicion on perceived communication and nonverbal behavior dynamics. Human Communication Research, 22, 163-196.

34. Burgoon, J.K., Buller, D.B., Ebesu, A.S., & Rockwell, P. (1994). Interpersonal deception: V. Accuracy in deception detection. Communication Monographs, 61, 303-325.

35. Burgoon, J.K., Buller, D.B., & Floyd, K. (2001). Does participation affect deception success? A test of the interactivity principle. Human Communication Research, 27, 503-533.

36. Burgoon, J. K., Buller, D. B., Floyd, K., & Grandpre, J. (1996). Deceptive realities: Sender, receiver, and observer perspectives in deceptive conversations. Communication Research, 23, 724-748.

37. Burgoon, J.K., Buller, D.B., Guerrero, L.K., Afifi, W.A., & Feldman, C.M. (1996). Interpersonal deception: XII. Information management dimensions underlying deceptive and truthful messages. Communication Monographs, 63, 50-69.

38. Burgoon, J.K., Buller, D.B., White, C.H., Afifi, W., & Buslig, A.L.S. (1999). The role of conversational involvement in deceptive interpersonal interactions. Personality and Social Psychology Bulletin, 25, 669-685.

39. Burgoon, J.K., Stoner, G.M., Bonito, J.A., & Dunbar, N.E. (2003). Trust and deception in mediated communication. Paper presented at the 36^th Hawaii conference on system sciences.

40. Cardena, E. (1983). The face of deception: Perception of lying and facial expression. Unpublished M.A. thesis, York University.

41. Cheng, H.C. (1996). A study of baseline familiarity on deception detection: Effects of rehearsal and consequences for lying. Unpublished MA thesis, SUNY-Buffalo Dept. of Communication.

42. Childers, C.D. (1980). An exploration of multiple channel evaulations in attributions of deception. Unpublished M.A. thesis, University of Mississippi.

43. Christensen, D. (1980). Decoding of intended versus unintended nonverbal messages as a function of social skill and anxiety. Unpublished doctoral dissertation, University of Connecticut.

44. Clark, L.M. (1983). Training humans to become better decoders of deception. Unpublished M.A. thesis, University of Georgia.

45. Cogburn, R.K. (1993). A study of psychopathy and its relation to success in interpersonal deception. Unpublished doctoral dissertation, University of Oregon.

46. Dawes, D. H. (1988). Preconscious processing and the detection of deception, Unpublished Ph.D. dissertation, University of Texas.

47. DePaulo, B.M., Blank, A.L., Swain, G.W., & Hairfield, J.G. (1992). Expressiveness and expressive control. Personality and Social Psychology Bulletin, 18, 276-285.

48. DePaulo, B. M., Jordan, A., Irvine, A., & Laser, P. S. (1982). Age changes in the detection of deception. Child Development, 53, 701‑709.

49. DePaulo, B.M., Kirkendol, S.E., Tang, J., & O'Brien, T.P. (1988). The motivational impairment effect in the communication of deception: Replications and extensions. Journal of Nonverbal Behavior, 12, 177-202.

50. DePaulo, B. M., Lanier, K., & Davis, T. (1983). Detecting the deceit of the motivated liar. Journal of Personality and Social Psychology, 45, 1096‑1103.

51. DePaulo, B. M., Lassiter, G. D., & Stone, J. I. (1982). Attentional determinants of success at detecting deception and truth. Personality and Social Psychology Bulletin, 8, 273‑279.

52. DePaulo, B.M., LeMay, C.S., & Epstein, J.A. (1991). Effects of importance of success and expectations for success on effectiveness at deceiving. Personality and Social Psychology Bulletin, 17, 14-24.

53. DePaulo, B.M., & Pfeifer, R.L. (1986). On‑the‑job experience and skill at detecting deception. Journal of Applied Social Psychology, 16, 249‑267.

54. DePaulo, B. M., & Rosenthal, R. (1979). Telling lies. Journal of Personality and Social Psychology, 37, 1713‑1722.

55. DePaulo, B.M., Rosenthal, R., Green, C. R., & Rosenkrantz, J. (1982). Diagnosing deceptive and mixed messages from verbal and nonverbal cues. Journal of Experimental Social Psychology, 18, 433‑446.

56. DePaulo, B.M., Stone, J.I., & Lassiter, G.D. (1985). Telling ingratiating lies: Effects of target sex and target attractiveness on verbal and nonverbal deceptive success. Journal of Personality and Social Psychology, 48, 1191‑1203.

57. DePaulo, B.M., & Tang, J. (1994). Social anxiety and social judgment: The example of detecting deception. Journal of Research in Personality, 28, 142-153.

58. DePaulo, B.M., Tang, J., & Stone, J.I. (1987). Physical attractiveness and skill at detecting deception. Personality and Social Psychology Bulletin, 13, 177-187.

59. DePaulo, P.J., & DePaulo, B.M. (1989). Can deception by salespersons and customers be detected through nonverbal behavioral cues? Journal of Applied Social Psychology, 19, 1552-1577.

60. deTurck, M.A. (1991). Training observers to detect spontaneous deception. Effects of gender. Communication Reports, 4, 81-89.

61. deTurck, M.A., Feeley, T.H., & Anastasiou, L. (1997). Effects of motivation to deceive and rehearsal on deception detection. Unpublished manuscript.

62. deTurck, M.A., Feeley, T.H., & Roman, L.A. (1997). Vocal and visual cue training in behavioral lie detection. Communication Research Reports, 14, 249-259.

63. deTurck, M.A., Harszlak, J.J., Bodhorn, D.J., & Texter, L.A. (1990). The effects of training social perceivers to detect deception from behavioral cues. Communication Quarterly, 38, 189-199.

64. deTurck, M.A., & Miller, G.R. (1990). Training observers to detect deception: Effects of self-monitoring and rehearsal. Human Communication Research, 16, 603-620.

65. Ekman, P., Frank, M., & O’Sullivan, M. (1994). Detecting deceit from demeanor. Unpublished manuscript. University of California at San Francisco.

66. Ekman, P., & Friesen, W.V. (1974). Detecting deception from the body or face. Journal of Personality and Social Psychology, 29, 288-298.

67. Ekman, P., Friesen, W.V., & Simons, R.C. (1985). Is the startle reaction an emotion? Journal of Personality and Social Psychology, 49, 1416-1426.

68. Ekman, P., & O'Sullivan, M. (1991). Who can catch a liar? American Psychologist, 46, 913-920.

69. Ekman, P., O’Sullivan, M., & Frank, M.G. (1999). A few can catch a liar. Psychological Science, 10, 263-266.

70. Elaad, E. (2003). Effects of feedback on the overestimated capacity to detect lies and the underestimated ability to tell lies. Applied Cognitive Psychology, 17, 249-263.

71. Elliott, G.L. (1979). Some effects of deception and level of self-monitoring on planning and reacting to a self-presentation. Journal of Personality and Social Psychology, 37, 1282-1292.

72. Etcoff, N.L., Ekman, P., Magee, J.J., & Frank, M.G. (2000). Lie detection and language comprehension. Nature, 405, 139.

73. Fan, R.M., Wagner, H.L., & Manstead, A.S.R. (1995). Anchoring, familiarity, and confidence in the detection of deception. Basic and Applied Social Psychology, 17, 83-96.

74. Fay, P.J., & Middleton, W.C. (1941). The ability to judge truth-telling, or lying, from the voice as transmitted over a public address system. Journal of Genetic Psychology, 24, 211-215.

75. Feeley, T.H. (1996). Conversational competence and perceptions of honesty in interpersonal deception. Paper presented at the annual meeting of the Speech Communication Association, San Diego, CA.

76. Feeley, T.H., & deTurck, M.A. (1995). Global cue usage in behavioral lie detection. Communication Quarterly, 43, 420-430.

77. Feeley, T.H., & deTurck, M.A. (1997a). Case-relevant vs case-irrelevant questioning in experimental lie detection. Communication Reports, 10, 35-45.

78. Feeley, T.H., deTurck, M.A. (1997b). Perceptions of communication as seen by the actor and as seen by the observer: The case of lie detection. Paper presented at the International Communication Association, Montreal, Canada.

79. Feeley, T.H., deTurck, M.A., & Young, M.J. (1995). Baseline familiarity in lie detection. Communication Research Reports, 12, 160-169.

80. Ferran-Urdaneta, C., & Storck, J. (1997) Truth or deception: The impact of videoconferencing for job interviews. Proceedings of the 18^th International Conference on Information Systems. Atlanta, GA.

81. Fiedler, K., Schmid, J., Kurzenhauser, S., & Schroter, V. (1997). Lie detection as an attribution process: The anchoring effect revisited. Unpublished manuscript.

82. Fiedler, K., & Walka, I. (1993). Training lie detectors to use nonverbal cues instead of global heuristics. Human Communication Research, 20, 199-223.

83. Fontenot, K.A. (1993). The relationship of conversational sensitivity and employment interview experience to deception detection in employment interviews. Unpublished doctoral dissertation, Louisiana State University.

84. Forrest, J.A. (2001). A social cognitive model of detecting deception. Unpublished Ph.D. dissertation. University of Massachusetts.

85. Forrest, J.A., & Feldman, R.S. (2000). Detecting deception and judge's involvement: Lower task involvement leads to better lie detection. Personality and Social Psychology Bulletin, 26, 118-125.

86. Fraidin, S.N., Hollingshead, A.B., & Kruger, J. (2002). Effects of suspicion on perceptions, judgments, and truth bias in lie detection. Paper presented at the annual meeting of the Midwestern Psychological Association, Chicago.

87. Frank, M.G. (1989). Human lie detection ability as a function of the liar's motivation. Unpublished doctoral dissertation, Cornell University.

88. Frank, M.G., & Ekman, P. (1997). The ability to detect deceit generalizes across different types of high-stakes lies. Journal of Personality and Social Psychology, 72, 1429-1439.

89. Frank, M.G., Paolantonio, N., Feeley, T.H., & Servoss, T.J. (2004). Individual and small group accuracy in judging truthful and deceptive communication. Group Decision and Negotiation, 13, 45-59.

90. Fugita, S.S., Hogrebe, M.C., & Wexley, K.N. (1980). Perceived expertise in detecting deception, successfulness of deception and nonverbal cues. Personality and Social Psychology Bulletin, 6, 637-643.

91. Galarza, L. (1996). On the accuracy of detecting deception in selection interviews: The effects of applicant rehearsal, applicant job interest, and self-monitoring. Unpublished M.A. thesis, Rice University.

92. Garrido, E., & Masip J. (2001). Previous exposure to the sender’s behavior and accuracy at judging credibility. In R. Roesch, R.R. Corrado, & Dempster, R. (Eds.) Psychology in the courts (pp. 271-287). London: Routledge.

93. Geis, F.L., & Moon, T.H. (1981). Machiavellianism and deception. Journal of Personality and Social Psychology, 41, 766-775.

94. Geizer, R.S., Rarick, D.L., & Soldow, G.F. (1977). Deception and judgment accuracy: A study in person perception. Personality and Social Psychology Bulletin, 3, 446-449.

95. George, J.F., Marett, K., Burgoon, J.K., Crews, J., Cao, J., Lin, M., & Biros, D.P. (2004). Training to detect deception: An experimental investigation. Proceedings of the 37^th Hawaii International Conference on System Sciences. Big Island, HI.

96. Gilovich, T., Savitsky, K., & Medvec, V.H. (1998). The illusion of transparency: Biased assessments of others’ ability to read one’s emotional states. Journal of Personality and Social Psychology, 75, 332-346.

97. Goldenberg, E.I. (1996). Emotion and perceived transparency in the context of lying versus honesty. Unpublished honours thesis, University of Manitoba.

98. Granhag, P.A., & Strömwall, L.A. (2001). Deception detection based on repeated interrogations. Legal & Criminological Psychology, 6, 85-101.

99. Hall, S. (1989). The generalizability of learning to detect deception in effective and ineffective deceivers. Unpublished Ph.D. thesis, Auburn University.

100. Harrison, A.A., Hwalek, M., Raney, D., & Fritz, J.G. (1978). Cues to deception in an interview situation. Social Psychology, 41, 156-161.

101. Hartwig, M., Granhag, P.A., Stromwall, L.A., & Andersson, L.O. (2002). Suspicious minds: Criminals’ ability to detect deception. Paper presented at 12^th European Conference on Psychology and Law: Leuven, Belgium: September 2002.

102. Hartwig, M., Granhag, P.A., Strömwall, L.A., & Vrij, A. (2002). Deception detection: Effects of conversational involvement and probing. Göteborg Psychological Reports, 32, 1-12.

103. Hartwig, M., Granhag, P.A., Strömwall, L.A., & Vrij, A. (2004). Police officers’ lie detection accuracy: Interrogating freely versus observing video. Police Quarterly, 7, 429-456.

104. Heinrich, C.U., & Borkenau, P. (1998). Deception and deception detection: The role of cross-modal inconsistency. Journal of Personality, 66, 687-711.

105. Hemsley, G.D. (1977). Experimental studies in the behavioral indicants of deception. Unpublished doctoral dissertation, University of Toronto.

106. Hendershot, J. (1981). Detection of deception in low and high socialization subjects with trained and untrained judges. Unpublished M.S. thesis, Auburn University.

107. Hess, U., & Kleck, R.E. (1994). The cues decoders use in attempting to differentiate emotion-elicited and posed facial expressions. European Journal of Social Psychology, 24, 367-381.

108. Hocking, J.E., Bauchner, J., Kaminski, E.P., & Miller, G.R. (1979). Detecting deceptive communication from verbal, visual, and paralinguistic cues. Human Communication Research, 6, 33-46.

109. Horn, D.B. (2001). Seeing is believing: Video quality and lie detection. Unpublished Ph.D. dissertation, University of Michigan.

110. Hubbell, A.P., Mitchell, M.M., & Gee, J.C. (2001). The relative effects of timing of suspicion and outcome involvement on biased message processing. Communication Monographs, 68, 115-132.

111. Hughes, S.D. (1997). The truth ‘bias’ in judging acquaintances. Unpublished M.S. thesis, Texas Christian University.

112. Hurd, K., & Noller, P. (1988). Decoding deception: A look at the process. Journal of Nonverbal Behavior, 12, 217-233.

113. Johnson, A.K, et al (2005). Me, myself, and lie: The role of self-awareness in deception. Personality and Individual Differences, 38, 1847-1853.

114. Kassin, S.M., & Fong, C.T. (1999). “I’m innocent!”: Effects of training on judgments of truth and deception in the interrogation room. Law & Human Behavior, 23, 499-516.

115. Kassin, S.A., Meissner, C.A., & Norwick, R.J. (2002). “I'd know a false confession if I saw one": A comparative study of college students and police investigators. Law and Human Behavior, 29, 211-227.

116. Kirk, C.T. (1993). A nonspecific factor in deception judgments. Unpublished M.S. thesis, Texas Christian University.

117. Koehnken, G. (1987). Training police officers to detect deceptive eyewitness statements: Does it work? Social Behaviour, 2, 1-17.

118. Kraut, R.E. (1978). Verbal and nonverbal cues in the perception of lying. Journal of Personality and Social Psychology, 36, 380-391.

119. Kraut, R.E., & Poe, D. (1980). Behavioral roots of person perception: The deception judgments of customs inspectors and laymen. Journal of Personality and Social Psychology, 39, 784-798.

120. Kurasawa, T. (1988). Effects of contextual expectancies on deception-detection. Japanese Psychological Research, 30, 114-121.

121. Landry, K.L., & Brigham, J.C. (1992). The effect of training in Criteria-Based Content Analysis on the ability to detect deception in adults. Law and Human Behavior, 16, 663-676.

122. Lane, J.D., & DePaulo, B.M. (1999). Completing Coyne's cycle: Dysphorics' ability to detect deception. Journal of Research in Personality, 33, 311-329.

123. Lashley, B.R. (1993). Anxiety cues versus verbal-nonverbal inconsistency: Two factors in deception judgments. Unpublished M.S. thesis, Texas Christian University.

124. Lavrakas, P.J., & Maier, R.A. (1979). Differences in human ability to judge veracity from the audio medium. Journal of Research in Personality, 13, 139-153.

125. Lee, C.C. (2000). The effect of levels of probing on the detection of deception in audit oral evidence. Unpublished doctoral dissertation, Southern Illinois University.

126. Lee, C.C. & Welker, R.B. (2002). Professional skepticism and the detection of intentional oral misrepresentation. Working paper, Florida International University.

127. Levine, T.R., Anders, L.N., et al (2000). Norms, expectations, and deception: A norm violation model of veracity judgments. Communication Monographs, 67, 123-137.

128. Levine, T.R., & McCornack, S.A. (2001). Behavioral adaptation, confidence, and heuristic-based explanations of the probing effect. Human Communication Research, 27, 471-502.

129. Levine, T.R., Park, H.S., & McCornack, S.A. (1999). Accuracy in detecting truths and lies: Documenting the "veracity effect". Communication Monographs, 66, 125-144.

130. Lin , Y.C. (1999). A study of training on deception detection: The effects of the specific six cues versus heuristics on deception detection accuracy. Unpublished M.A. thesis, SUNY-Buffalo.

131. Littlepage, G.E., McKinnie, R., & Pineault, M.A. (1983). Relationship between nonverbal sensitivities and detection of deception. Perceptual and Motor Skills, 57, 651-657.

132. Littlepage, G.E., Maddox, J., & Pineault, M.A. (1985). Recognition of discrepant nonverbal messages and detection of deception. Perceptual and Motor Skills, 60, 119-124.

133. Littlepage, G.E., & Pineault, M.A. (1979). Detection of deceptive factual statements from the body and the face. Personality and Social Psychology Bulletin, 5, 325-328.

134. Littlepage, G.E., & Pineault, M.A. (1981). Detection of truthful and deceptive interpersonal communications across information transmission modes. Journal of Social Psychology, 114, 57-68.

135. Littlepage, G.E., & Pineault, M.A. (1985). Detection of deception of planned versus spontaneous communications. Journal of Social Psychology, 125, 195-201.

136. Littlepage, G.E., & Pineault, T. (1978). Verbal, facial, and paralinguistic cues to the detection of truth and lying. Personality and Social Psychology Bulletin, 4, 461-464.

137. Littlepage, G.E., Tang, D.W., & Pineault, M.A. (1986). Nonverbal and content factors in the detection of deception in planned and spontaneous communications. Journal of Social Behavior and Personality, 1, 439-450.

138. Lusby, D. J. (1999). Effects of motivation to lie and sanctioned deception of the accuracy of observers’ veracity judgments. Unpublished M.A. thesis, SUNY-Buffalo.

139. Malcolm, S.R., & Keenan, J.P. (2005). Hemispheric asymmetry and deception detection. Laterality, 10, 103-110.

140. Mann, S., Vrij, A., & Bull, R. (2004). Detecting true lies: Police officers’ ability to detect suspects’ lies. Journal of Applied Psychology, 89, 137-149.

141. Manstead, A.S., Wagner, H.L., & MacDonald, C.J. (1984). Face, body, and speech as channels of communication in the detection of deception. Basic and Applied Social Psychology, 5, 317-332.

142. Manstead, A.S.R., Wagner, H.L., & MacDonald, C.J. (1986). Deceptive and nondeceptive communications: Sending experience, modality, and individual abilities. Journal of Nonverbal Behavior, 10, 147-167.

143. Masip, J. (2002). Credibility assessment of the testimony on the basis of behavioral indicators in criminal justice settings. Unpublished doctoral dissertation, University of Salamanca.

144. .Masip, J., Garrido, E., & Herrero, C. (In press). Observers’ decision moment in deception detection experiments: Its impact on judgment, accuracy and confidence. International Journal of Psychology.

145. Masip, J., Garrido, E., & Herrero, C. (2003). Statement length and credibility judgments: Questioning the truth bias. Paper presented at the Psychology and Law International Interdisciplinary Conference, Edinburgh.

146. Meissner, C.A., & Kassin, S.M. (2002). “He’s guilty:” Investigator bias and judgments of truth and deception. Law and Human Behavior, 26, 469-480.

147. Millar, M., & Millar, K. (1995). Detection of deception in familiar and unfamiliar persons: The effects of information restriction. Journal of Nonverbal Behavior, 19, 69-84.

148. Millar, M.G., & Millar, K. (1997a). Effects of situational variables on judgments about deception and detection accuracy. Basic and Applied Social Psychology, 19, 401-410.

149. Millar, M.G., & Millar, K.U. (1997b). The effects of cognitive capacity and suspicion on truth bias. Communication Research, 24, 556-570.

150. Miller, G.R., & Fontes, N.E. (1979). Videotape on trial: A view from the jury box. Beverly Hills, CA: Sage.

151. Miller, G.R., deTurk, M.A., & Kalbfleisch, P.J. (1983). Self-monitoring, rehearsal, and deceptive communication. Human Communication Research, 10, 97-117.

152. Morris, K.A. (2003). Teaching students about classic findings on the detection of deception. Teaching of Psychology, 30, 111-113.

153. Motley, M.T. (1974). Acoustic correlates of lies. Western Speech, 38, 81-87.

154. Murray, J. (1983). The detection of interviewees' verbal deceptions from their accompanying overt nonverbal behavior. Unpublished doctoral dissertation, Florida State University.

155. Nance, C.H. (1993). The effect of specificity of experience on the ability to detect deception in same- and other-race target persons. Unpublished M.S. thesis, Florida State University.

156. Nance, C.H. (1995). The effects of high versus low context congruity on accuracy in detecting deception. Unpublished Ph.D. dissertation, Florida State University.

157. Oberleitner, D., McLarney-Vesotski, A., Bernieri, F., & Okdie, B. (2004). Musical mood induction and its effects on lie detection accuracy. Paper presented at the annual meeting of the Society for Personality and Social Psychology, Austin, TX.

158. Oldfield, S.E. (1999). The influence of self-monitoring on the detection of deception. Unpublished MA thesis, Radford University.

159. O’Sullivan, M. (2003). The fundamental attribution error in detecting deception: The boy-who-cried-wolf effect. Personality and Social Psychology Bulletin, 29, 1316-1327.

160. O'Sullivan, M., Ekman, P., & Friesen, W.V. (1988). The effect of comparisons on detecting deceit. Journal of Nonverbal Behavior, 12, 203-215.

161. Park, E.S., Levine, T.R., Harms, C.M., & Ferrara, M.H. (2002). Group and individual accuracy in deception detection. Communication Research Reports, 19, 99-106.

162. Parker, R.J. (1978). Age, sex, and the detection of deception through nonverbal cues. Unpublished doctoral dissertation, California School of Professional Psychology.

163. Pietras, K.R., & Bond, C.F., Jr. (1985). The reliability of lie detection judgments. Paper presented at the annual meeting of the Eastern Psychological Association, Boston, MA.

164. Porter, S., Campbell, M.A., Stapleton, J., & Birt, A.R. (2002). The influence of judge, target, and stimulus characteristics on the accuracy of detecting deceit. Canadian Journal of Behavioural Science, 34, 172-185.

165. Porter, S., Woodworth, M., & Birt, A.R. (2000). Truth, lies, and videotape: An investigation of the ability of federal parole officers to detect deception. Law & Human Behavior, 24, 643-658.

166. Potamkin, G.G. (1982). Heroin addicts and nonaddicts: The use and detection of nonverbal deception cues. Unpublished doctoral dissertation, California School of Professional Psychology.

167. Raichle, W. (1990). The roles of verbal and nonverbal behavior in impression formation during an employment interview. Unpublished doctoral dissertation, New York University.

168. Riggio, R.E., & Friedman, H.S. (1983). Individual differences and cues to deception. Journal of Personality and Social Psychology, 45, 899-915.

169. Riggio, R.E., Tucker, J., & Throckmorton, B. (1987). Social skills and deception ability. Personality and Social Psychology Bulletin, 13, 568-577.

170. Rockwell, P. (1996). Hemispheric differences in detection of deception with content-filtered speech. Perceptual and Motor Skills, 82, 1241-1242.

171. Rotkin, H.G. (1980). Information used in detecting deception. Unpublished doctoral dissertation, New York University.

172. Rudin, J.P. (1999). Effect of reward size in performance of a deception detection task. Proceedings of the annual meeting of the Southwest Academy of Management (pp. 182-184). Houston.

173. Rudin, J.P., & Wazeter, D. (1998). Employment interviewers as lie detectors: More harm than good? Central Business Review, 17, 32-36.

174. Russo-Devosa, Y.M. (1999). The credibility of adult and child witnesses: Implications for detecting deception in child sexual abuse cases. Unpublished MS thesis, Pennsylvania State University.

175. Sahlman, J.M., & Koper, R.J. (1992). Do you hear what I hear? Deception detection by the blind. Paper presented at the annual meeting of the International Communication Association, Miami: May 1992.

176. Sakai, D.J. (1981). Nonverbal communication in the detection of deception among women and men. Unpublished doctoral dissertation, University of California, Davis.

177. Schoephoerster, B.T. (1996). Deception detection accuracy: The effects of suspicion and antisocial personality traits. Unpublished MA thesis, University of Nevada at Las Vegas.

178. Seager, P.B. (2001). Improving the ability of people to detect lies. Unpublished doctoral dissertation, University of Hertfordshire.

179. Seiter, J. S. (1997). Honest or deceitful? A study of persons’ mental models for judging veracity. Human Communication Research, 24, 216-259.

180. Skaggs, L.M. (1993). The effects of behavior coordination on judgments of deception. Unpublished M.S. thesis, Texas Christian University.

181. Steinberg, J.R., & Bernieri, F. (2001). How locus of control affects accuracy of detecting deception. Paper presented at the annual meeting of the American Psychological Society, Toronto, CA.

182. Stiff, J.B., & Miller, G.R. (1986). "Come to think of it...": Interrogative probes, deceptive communication, and deception detection. Human Communication Research, 12, 339-357.

183. Streeter, L.A., Krauss, R.M., Geller, V., Olsen, C., & Apple, W. (1977). Pitch changes during attempted deception. Journal of Personality and Social Psychology, 35, 345-350.

184. Strömwall, L.A., Granhag, P.A., & Jonsson, A.C. (2003). Deception among pairs: “Let’s say we had lunch and hope they will swallow it!” Psychology, Crime, and Law, 9, 109-124.

185. Swann, W.B., Jr., Silvera, D.H., & Proske, C.U. (1995). On "knowing your partner": Dangerous illusions in the age of AIDS? Personal Relationships, 2, 173-186.

186. Swinkels, A.H. (1989). The effects of cognitive busyness on human lie detection ability. Unpublished Ph.D. dissertation, University of Texas at Austin.

187. Taylor, R. (1999) Relationships between accuracy of detecting deception and beliefs about the cues involved. Paper presented at the Division of Forensic Psychology Annual Conference. Cambridge.

188. Toris, C., & DePaulo, B.M. (1984). Effects of actual deception and suspiciousness of deception on interpersonal perceptions. Journal of Personality and Social Psychology, 47, 1063‑1073.

189. Tornqvist, J.S. (2002) The detectability and believability of lies and truths in the e-mail medium. Unpublished Ph.D. dissertation, University of Virginia.

190. Vrij, A. (1993). Credibility judgments of detectives: The impact of nonverbal behavior, social skills, and physical characteristics on impression formation. Journal of Social Psychology, 133, 601-610.

191. Vrij, A. (1994). The impact of information and setting on detection of deception by police detectives. Journal of Nonverbal Behavior, 18, 117-136.

192. Vrij, A., & Baxter, M. (1999). Accuracy and confidence in detecting truths and lies in elaborations and denials: Truth bias, lie bias and individual differences. Expert Evidence, 7, 25-36.

193. Vrij, A., Edward, K., & Bull, R. (2001). Police officers’ ability to detect deceit: The benefit of indirect deception detection measures. Legal & Criminological Psychology, 6, 185-196.

194. Vrij, A., & Graham, S. (1997). Individual differences between liars and the ability to detect lies. Expert Evidence, 5, 144-148.

195. Vrij, A., Harden, F., Terry, J., Edward, K., & Bull, R. (2001). The influence of personal characteristics, stakes and lie complexity on the accuracy and confidence to detect deceit. In R. Roesch, R.R. Corrado, and R. Dempster (Eds.) Psychology in the courts (pp. 289-302). London: Routledge.

196. Wetzel, C., Addison, S., & Mueller, J. (2005). Detecting imposters: To tell the truth, we don’t do it well. Unpublished manuscript, Rhodes College.

197. Williams, D. (1989). Effects of familiarity and communication mode on the detection of deception. Unpublished M.A. thesis, University of West Florida.

198. Wilson, S.J. (1975). Channel differences in the detection of deception. Unpublished doctoral dissertation, Florida State University.

199. Yang, C.C. (1996). The effects of training, rehearsal, and consequences for lying on deception detection accuracy. Unpublished M.A. thesis, SUNY-Buffalo.

200. Zuckerman, M., Amidon, M.D., Bishop, S.E., & Pomerantz, S.D. (1982). Face and tone of voice in the communication of deception. Journal of Personality and Social Psychology, 43, 347-357.

201. Zuckerman, M., DeFrank, R.S., Hall, J.A., Larrance, D.T., & Rosenthal, R. (1979). Facial and vocal cues of deception and honesty. Journal of Experimental Social Psychology, 15, 378-396.

202. Zuckerman, M., Driver, R., & Guadagno, N.S. (1985). Effects of segmentation patterns on the perception of deception. Journal of Nonverbal Behavior, 9, 160-168.

203. Zuckerman, M., Fischer, S.A., Osmun, R.W., Winkler, B.A., & Wolfson, L.R. (1987). Anchoring in lie detection revisited. Journal of Nonverbal Behavior, 11, 4-12.

204. Zuckerman, M., Koestner, R., & Alton, A.O. (1984). Learning to detect deception. Journal of Personality and Social Psychology, 46, 519-528.

205. Zuckerman, M., Koestner, R., & Colella, M.J. (1985). Learning to detect deception from three communication channels. Journal of Nonverbal Behavior, 9, 188-194.

206. Zuckerman, M., Koestner, R., Colella, M.J., & Alton, A.O. (1984). Anchoring in the detection of deception and leakage. Journal of Personality and Social Psychology, 47, 301-311.

Appendix B

Coding of the 206 Documents

Doc# Send# Rec# Truth% Acc% d Med Mot Prep Base Inter Exp

1 72 72 52.31 2 0 0 0 0 0

1 72 72 48.88 1 0 0 0 0 0

1 72 72 54.22 3 0 0 0 0 0

2 200 200 51.50 58.50 3 0 0 0 0 0

3 52 8 50.60 0.414 3 0 0 0 2 0

4 8 69 50.70 60.50 3 1 1 0 2 0

5 8 24 47.96 53.62 1 0 1 0 2 0

6 30 25 58.94 59.07 3 0 1 0 2 0

6 30 25 56.18 58.33 3 0 1 0 2 0

7 12 60 45.48 4 1 0 0 2 0

8 8 30 52.81 1 0 0 0 2 0

9 60 150 56.83 0.602 3 1 1 0 0 0

10 10 25 67.50 3 1 0 1 2 0

11 32 20 48.60 52.76 2 0 0 0 0 0

11 32 20 52.17 51.02 3 0 0 0 0 0

11 32 20 49.53 51.51 1 0 0 0 0 0

11 32 40 60.31 57.27 2 0 0 0 0 0

11 32 40 60.14 53.60 3 0 0 0 0 0

11 32 40 51.85 48.92 1 0 0 0 0 0

11 32 40 56.54 51.08 1 2 0 0 0 0

11 32 40 57.02 49.53 2 2 0 0 0 0

11 32 40 57.49 54.35 3 2 0 0 0 0

12 32 2 53.00 56.00 3 0 0 0 0 0

12 32 45 49.63 49.79 3 0 0 0 0 0

13 32 34 49.26 63.46 3 1 0 0 2 0

14 48 40 57.39 53.54 3 0 0 0 0 0

15 24 48 57.17 51.97 1 0 0 0 0 0

15 24 48 54.92 54.80 1 0 0 0 0 0

16 48 40 67.42 55.15 3 0 0 0 0 0

16 48 192 56.08 50.12 1 0 0 0 0 0

17 48 67 57.40 47.70 5 0 0 0 0 0

18 16 15 57.62 60.42 3 1 0 0 0 0

19 48 40 52.58 46.85 1 0 0 0 0 0

19 48 50 52.93 48.60 2 0 0 0 0 0

19 48 40 55.00 48.13 3 0 0 0 0 0

19 48 196 55.64 48.57 3 0 0 0 0 0

19 32 14 54.69 51.56 3 0 0 0 0 0

20 12 12 45.85 48.63 3 0 1 0 1 1

20 15 15 42.20 58.90 3 0 1 0 1 1

20 15 15 76.65 53.35 3 0 1 0 1 0

21 36 36 35.44 49.98 3 0 0 0 1 1

21 36 36 61.48 58.68 3 0 0 0 1 0

22 56 56 53.00 60.00 3 0 0 0 1 1

22 56 56 74.75 51.50 3 0 0 0 1 0

23 48 31 46.64 53.00 3 1 1 0 0 0

23 48 31 49.98 3 1 1 0 0 0

23 48 33 46.72 50.59 3 1 1 0 2 0

23 48 31 50.13 3 1 1 0 2 0

24 15 48 52.26 53.46 3 0 0 0 0 0

25 4 50 42.00 3 1 0 0 2 0

26 16 50 44.50 3 1 0 0 2 0

27 16 66 41.65 -0.379 3 1 0 0 2 0

28 40 20 1.410 3 0 0 0 1 1

29 8 124 59.59 57.09 3 0 1 0 2 1

30 2 92 52.00 60.00 3 0 0 0 2 0

31 2 50 76.00 55.00 3 0 0 0 2 0

31 2 150 60.00 59.00 3 0 0 0 2 0

32 17 17 82.35 45.14 -0.471 3 1 0 0 2 0

32 17 17 70.59 57.64 0.876 2 1 0 0 2 0

32 17 17 47.06 35.42 0.013 5 1 0 0 2 0

33 60 60 0.365 3 0 0 1 1 0

34 18 18 0.646 3 0 1 0 1 0

34 19 19 0.399 3 0 1 0 1 1

35 16 16 0.278 3 0 0 1 1 0

36 13 13 0.620 3 0 0 0 1 0

36 10 10 0.000 3 0 0 0 2 0

37 106 2 1.195 3 0 4 0 2 0

38 61 61 0.498 3 0 1 0 1 0

39 16 16 0.099 3 0 0 0 1 0

39 16 16 0.796 3 0 0 0 2 0

39 16 16 0.695 2 0 0 0 2 0

39 16 16 -0.453 5 0 0 0 2 0

40 10 36 52.04 52.58 1 0 0 0 2 0

41 12 100 54.00 3 2 4 2 2 0

42 5 16 52.49 71.27 3 0 0 0 2 1

42 5 122 54.98 60.24 3 0 0 0 2 0

42 5 151 54.62 71.00 3 0 0 0 2 0

43 12 38 59.48 59.42 3 0 0 0 2 0

43 12 46 57.82 58.83 3 0 0 0 2 0

43 12 19 70.91 48.24 3 0 0 0 2 0

44 20 46 54.89 67.49 3 0 0 1 2 0

45 52 9 0.120 3 1 0 0 2 0

46 8 20 67.88 56.63 3 0 0 0 0 0

47 32 115 -0.060 1 1 1 0 0 0

47 32 115 0.110 5 1 1 0 0 0

47 32 115 0.230 2 1 1 0 0 0

47 32 115 0.240 3 1 1 0 0 0

48 8 43 0.311 3 0 1 0 0 0

49 131 74 0.361 3 1 1 0 0 0

50 32 24 0.620 3 1 4 0 1 0

50 32 64 0.583 1 1 4 0 2 0

51 12 11 -0.258 3 0 1 0 0 0

52 96 3 0.307 3 1 1 0 1 0

52 96 816 0.375 5 1 1 0 2 0

53 16 114 60.20 52.30 2 1 4 0 2 1

53 16 144 61.00 52.90 2 1 4 0 2 1

53 16 161 62.00 54.30 2 1 4 0 2 0

54 40 40 0.810 3 0 1 0 0 0

55 40 16 0.040 1 0 1 0 0 0

55 40 16 -0.130 2 0 1 0 0 0

55 40 16 0.270 5 0 1 0 0 0

55 40 16 0.310 3 0 1 0 0 0

55 40 16 0.570 2 0 1 0 0 0

56 64 271 0.164 3 1 1 0 0 0

57 64 102 0.360 3 1 1 0 0 0

58 64 102 0.340 3 1 1 0 0 0

59 14 107 0.001 3 1 1 0 0 0

60 16 91 58.97 3 1 0 0 2 0

61 12 198 52.17 3 1 4 0 2 0

62 8 41 48.50 57.50 3 1 0 1 2 0

63 8 94 54.00 3 1 0 1 2 0

64 32 195 50.00 3 1 4 0 2 0

65 20 43 39.05 62.80 3 2 1 0 2 1

65 30 23 43.05 61.60 3 2 1 0 2 1

66 16 113 46.63 46.16 1 1 0 0 2 0

66 16 120 45.21 52.27 1 1 0 0 2 0

67 14 98 60.00 1 0 4 0 2 0

68 10 39 52.82 3 1 0 0 2 0

68 10 34 64.12 3 1 0 0 2 1

68 10 73 55.34 3 1 0 0 2 1

68 10 67 57.61 3 1 0 0 2 1

68 10 60 55.67 3 1 0 0 2 1

68 10 110 56.73 3 1 0 0 2 1

68 10 126 55.79 3 1 0 0 2 1

69 10 36 53.05 50.80 3 2 1 0 2 1

69 10 84 51.10 62.00 3 2 1 0 2 1

69 10 125 50.70 57.70 3 2 1 0 2 0

69 10 107 46.45 67.50 3 2 1 0 2 1

69 10 209 47.75 62.10 3 2 1 0 2 1

70 8 60 49.60 45.60 3 1 1 0 0 1

71 64 6 0.000 3 1 1 0 1 0

72 10 10 47.00 3 1 0 0 2 0

72 10 48 46.00 3 1 0 0 2 0

73 14 56 55.50 58.00 3 1 0 0 2 0

74 6 47 44.53 55.75 2 0 0 0 2 0

75 95 95 82.25 50.75 3 0 0 1 1 0

76 12 50 59.00 63.00 3 0 0 1 2 0

77 4 30 65.00 3 0 0 1 2 0

78 95 95 74.25 55.75 3 0 0 0 2 0

79 4 52 59.05 59.50 3 0 0 2 2 0

80 30 28 86.36 58.39 4 1 0 0 1 0

81 12 32 0.800 3 0 1 0 2 0

82 10 24 52.50 53.00 0.613 3 0 1 0 2 0

83 2 80 60.04 54.97 3 1 4 0 2 1

83 2 150 53.24 57.57 3 1 4 0 2 0

84 16 68 60.85 55.33 0.513 3 0 1 0 0 0

84 30 40 0.355 3 1 0 0 2 0

85 16 60 0.493 3 1 1 0 0 0

86 24 46 0.030 3 1 1 0 0 0

86 24 55 57.05 55.85 3 1 1 0 0 0

87 5 45 50.80 47.00 3 1 0 0 2 0

87 5 44 50.80 50.00 3 1 0 0 2 0

87 5 38 51.00 3 1 0 0 2 0

87 5 38 48.00 2 1 0 0 2 0

87 5 38 46.80 3 1 0 0 2 0

87 5 40 47.00 1 1 0 0 2 0

87 5 40 45.00 1 0 0 0 2 0

87 32 37 55.50 3 1 1 0 0 0

88 20 30 59.50 3 2 1 0 2 0

88 20 48 58.50 3 2 1 0 2 0

89 10 54 52.80 56.10 3 2 1 0 2 0

89 10 165 49.45 59.60 3 2 1 0 2 0

90 40 2 52.81 60.88 3 0 0 0 1 0

91 56 108 67.75 53.75 3 1 4 0 2 0

92 1 121 29.00 47.50 3 1 1 0 0 1

92 1 146 42.50 58.50 3 1 1 0 0 0

93 64 64 0.839 3 1 0 0 2 0

94 12 50 0.180 3 1 1 0 2 0

95 36 86 59.50 5 0 0 0 2 0

95 6 85 51.83 5 0 0 0 2 0

95 36 29 58.11 5 0 0 0 2 0

96 4 5 54.36 3 0 0 0 1 0

97 56 56 0.300 3 0 0 0 1 0

98 24 22 78.33 61.67 3 0 1 0 2 0

98 24 144 54.88 59.70 3 0 1 0 2 0

99 14 81 53.07 53.36 3 1 0 0 0 0

99 8 281 58.00 3 1 0 0 0 0

99 46 90 52.00 57.50 3 1 0 0 0 0

100 72 72 63.72 61.28 3 0 0 0 2 0

101 20 52 42.30 57.70 3 0 1 0 2 0

101 20 52 26.90 65.40 3 0 1 0 2 1

102 40 40 45.00 50.00 3 1 1 0 2 0

102 40 40 57.50 47.50 3 1 1 0 2 0

102 40 40 57.50 62.50 3 1 1 0 2 0

103 30 30 50.00 56.67 3 1 1 0 2 1

104 40 8 0.947 3 0 1 0 0 0

104 40 4 0.093 1 0 1 0 0 0

104 40 4 1.250 2 0 1 0 0 0

105 20 1 60.75 3 0 1 0 1 0

105 20 20 70.85 51.88 3 0 1 0 2 0

105 20 1 57.25 3 0 1 0 1 0

106 16 14 44.75 49.75 3 0 1 0 2 0

106 16 14 48.65 47.35 3 0 1 0 2 1

107 35 40 56.25 58.25 1.610 1 0 4 0 0 0

108 16 40 47.21 54.30 2 1 1 0 2 0

108 16 47 47.21 57.40 5 1 1 0 2 0

108 16 96 44.30 48.40 1 1 1 0 2 0

108 16 103 47.21 54.40 3 1 1 0 2 0

108 16 107 43.32 46.80 1 1 1 0 2 0

108 16 108 47.21 55.60 3 1 1 0 2 0

108 16 109 47.21 53.20 3 1 1 0 2 0

108 16 109 47.21 48.60 1 1 1 0 2 0

109 6 42 0.717 4 1 0 1 2 0

109 6 113 0.794 4 1 0 1 2 0

110 1 249 62.00 46.50 3 0 0 0 0 0

111 29 29 0.665 3 0 0 0 0 0

112 8 19 55.78 50.78 3 0 1 0 0 0

113 12 42 55.15 3 0 1 0 0 0

114 16 18 43.75 55.63 3 1 1 0 2 0

114 16 1 37.50 3 1 1 0 2 0

1156 10 48 58.65 45.35 3 0 0 0 2 1

1156 10 50 51.45 53.60 3 0 0 0 2 0

1156 10 29 65.50 54.50 2 0 0 0 2 1

1156 10 32 56.00 64.00 2 0 0 0 2 0

116 48 24 57.46 54.51 3 0 0 0 0 0

117 4 20 65.50 47.00 3 1 1 1 0 1

118 5 41 0.960 3 1 0 1 2 0

119 62 49 -0.510 3 1 0 0 2 0

119 62 39 -0.280 3 1 0 0 2 1

120 8 20 0.671 3 0 0 0 0 0

121 12 24 65.45 51.90 3 0 1 0 0 0

121 12 26 60.51 47.22 5 0 1 0 0 0

122 4 39 83.02 58.00 2 1 1 0 2 0

123 48 40 58.10 56.30 3 0 0 0 0 0

123 48 40 51.39 48.58 1 0 0 0 0 0

124 20 100 54.45 2 0 4 0 0 0

125 20 120 57.35 55.65 3 1 1 0 2 0

126 22 66 50.00 58.00 3 1 1 0 2 1

126 22 110 31.00 51.00 3 1 1 0 2 0

127 4 128 0.114 3 0 1 0 1 0

128 8 337 72.00 52.75 3 0 1 0 2 0

128 2 71 64.25 52.25 3 0 1 0 2 0

128 2 136 68.77 57.99 3 0 1 0 2 0

129 2 58 68.00 50.64 3 0 1 0 2 0

129 2 60 70.00 51.88 3 0 1 0 2 0

129 2 59 65.00 46.44 3 0 1 0 2 0

130 96 100 51.00 3 2 0 1 2 0

131 4 61 58.16 57.70 3 0 0 0 2 0

132 4 60 58.30 58.90 3 0 0 0 2 0

133 2 32 53.88 64.20 4 0 0 0 2 0

134 4 20 58.00 56.10 5 0 0 0 2 0

134 4 20 53.00 62.50 2 0 0 0 2 0

134 4 21 59.00 49.40 1 0 0 0 2 0

134 4 22 54.00 58.80 5 0 0 0 2 0

134 4 26 56.00 63.50 3 0 0 0 2 0

135 12 32 75.30 55.68 3 0 4 0 2 0

136 6 25 0.682 2 1 0 0 2 0

136 6 23 0.696 3 1 0 0 2 0

136 6 17 0.398 5 1 0 0 2 0

136 6 16 -0.213 1 1 0 0 2 0

137 4 26 74.28 51.70 2 0 4 0 2 0

137 4 36 80.21 51.40 5 0 4 0 2 0

137 4 36 60.94 52.92 1 0 4 0 2 0

138 24 24 31.00 3 1 0 1 2 0

138 24 72 53.67 3 0 0 1 2 0

139 2 32 64.62 58.87 2 0 1 0 0 0

140 14 99 46.54 64.89 3 1 3 0 2 1

141 4 43 0.314 3 0 1 0 0 0

141 2 43 -0.501 3 0 1 0 0 0

141 2 16 -0.484 3 0 1 0 0 0

141 2 16 0.693 3 0 1 0 0 0

141 4 15 -0.589 1 0 1 0 0 0

141 2 15 -1.033 1 0 1 0 0 0

142 36 36 0.315 4 0 1 0 0 0

143 24 54 54.63 54.32 3 1 1 0 0 0

144 1 224 27.48 51.48 2 1 1 0 0 1

145 24 52 52.57 55.13 3 1 1 0 0 0

146 16 44 35.50 47.50 3 1 1 0 2 1

147 31 53 50.00 4 0 1 0 0 0

147 27 58 54.50 4 0 1 0 0 0

148 8 242 70.00 49.00 3 1 0 0 2 0

149 12 107 86.00 49.00 3 1 0 0 0 0

150 16 94 53.23 3 1 1 0 2 0

150 16 99 53.43 3 1 1 0 2 0

151 32 151 51.00 3 1 4 1 2 0

152 82 82 44.84 3 0 1 0 1 0

153 17 10 48.79 2 0 0 0 2 0

154 16 48 48.39 59.04 1 0 0 0 2 0

155 16 189 50.25 0.630 3 0 1 0 0 0

156 8 150 62.00 47.00 0.032 3 1 1 0 2 0

156 8 147 55.50 39.50 -0.159 3 1 1 0 2 0

157 20 105 53.60 3 0 0 0 0 0

158 16 116 46.48 49.23 3 0 0 0 0 0

159 10 34 53.90 59.70 3 1 0 0 2 0

159 10 55 59.10 50.90 3 1 0 0 2 0

160 31 35 53.80 3 1 0 0 2 0

160 31 109 51.00 3 1 0 0 2 0

161 2 47 65.05 51.55 3 0 1 0 2 0

162 12 35 58.43 47.71 1 0 0 1 2 0

163 32 56 61.61 52.90 3 1 0 0 2 0

164 8 310 46.65 57.65 4 1 1 0 0 0

165 6 32 30.00 40.40 3 1 1 0 0 1

165 6 32 50.00 42.20 3 1 1 0 0 0

166 6 20 57.50 62.50 1 1 1 0 2 0

167 4 30 68.00 67.00 2 1 1 0 2 0

167 4 30 68.00 40.00 5 1 1 0 2 0

167 4 30 68.00 73.00 3 1 1 0 2 0

167 4 30 68.00 57.00 1 1 1 0 2 0

168 63 44 0.098 3 0 1 0 0 0

168 63 44 0.117 1 0 1 0 0 0

168 63 44 0.258 1 0 1 0 0 0

168 63 14 0.320 5 0 1 0 0 0

168 63 44 0.380 3 0 1 0 0 0

169 38 34 0.692 3 0 1 0 0 0

170 15 139 47.11 2 0 1 0 2 0

171 8 20 0.063 1 1 0 0 2 0

171 8 20 0.091 1 1 0 0 2 0

171 8 20 0.283 1 1 0 0 2 0

171 8 20 0.860 3 1 0 0 2 0

171 8 20 0.930 3 1 0 0 2 0

171 8 20 0.934 3 1 0 0 2 0

171 8 20 0.969 3 1 0 0 2 0

171 8 20 1.443 2 1 0 0 2 0

172 5 188 51.11 3 0 4 0 2 0

173 5 47 50.05 3 0 4 0 2 1

174 4 200 79.69 50.81 0.819 3 1 2 0 2 0

175 8 71 66.38 2 0 0 0 2 0

176 2 180 55.77 4 0 0 0 2 0

177 8 133 81.68 54.00 3 0 0 0 2 0

178 11 18 61.62 59.60 3 1 0 0 2 0

178 10 125 60.24 56.48 3 1 0 0 2 0

178 5 77 0.058 5 1 0 0 2 0

178 5 27 -0.392 2 0 0 0 2 0

178 5 25 54.00 39.20 -0.462 3 0 0 0 2 0

178 5 62 57.99 59.47 2 0 0 0 2 0

178 5 69 55.35 61.49 3 0 0 0 2 0

178 5 87 53.03 59.90 5 0 0 0 2 0

178 5 19 53.68 61.58 2 0 0 0 2 0

178 5 277 53.32 62.85 3 0 0 0 2 0

179 2 120 58.50 41.50 3 0 1 0 2 0

180 16 48 63.37 49.83 3 0 0 0 2 0

181 46 47 49.44 52.69 3 0 2 0 0 0

181 46 47 53.46 49.59 2 0 2 0 0 0

182 40 139 63.12 54.06 3 1 0 0 2 0

183 32 15 0.236 2 1 0 0 2 0

183 32 15 0.291 2 1 0 0 2 0

184 40 120 58.75 58.75 3 0 1 0 2 0

185 6 103 68.20 51.90 3 0 1 0 1 0

186 20 110 54.39 54.86 3 0 0 0 0 0

187 19 58 53.68 55.57 3 1 0 0 2 0

187 19 58 51.27 57.93 3 1 0 0 2 1

188 72 72 0.264 3 0 1 0 1 0

189 8 19 67.82 60.50 2 0 1 0 0 0

189 8 19 71.50 66.00 5 0 1 0 0 0

189 8 19 81.55 54.00 5 0 1 0 0 0

189 8 19 69.76 54.50 5 0 1 0 0 0

190 20 91 52.50 49.00 3 0 0 0 2 1

191 20 36 47.00 3 0 0 0 2 1

191 20 36 50.00 3 0 0 0 2 1

191 20 36 51.00 3 0 0 0 2 1

191 20 36 44.00 1 0 0 0 2 1

192 20 50 48.75 53.00 3 2 0 0 2 0

193 28 21 0.587 3 1 1 0 2 1

194 10 15 54.00 3 0 0 0 2 1

194 10 20 42.00 3 0 0 0 2 0

195 30 61 50.33 55.00 3 0 1 0 2 0

195 30 60 47.67 54.00 3 1 4 0 2 0

196 1 137 -0.216 3 1 1 0 1 0

196 1 548 0.189 3 1 1 0 1 0

197 2 40 86.21 51.59 3 1 0 2 0 0

197 2 40 90.00 52.50 2 1 0 2 0 0

198 10 36 62.00 1 0 0 0 2 0

198 10 36 60.50 1 0 0 0 2 0

198 10 36 57.20 3 0 0 0 2 0

198 10 36 53.20 2 0 0 0 2 0

198 10 36 48.70 1 0 0 0 2 0

199 4 22 56.50 3 0 2 1 2 0

199 4 44 53.50 3 1 3 1 2 0

200 60 9 0.639 4 0 0 0 0 0

201 60 77 54.20 51.80 0.200 2 0 1 0 0 0

201 60 60 46.70 52.70 1 0 1 0 0 0

202 24 68 0.186 1 0 1 0 0 0

202 24 68 0.299 3 0 1 0 0 0

203 8 201 1.250 3 0 1 0 0 0

204 8 19 55.00 3 0 1 0 0 0

204 8 43 58.00 3 0 1 0 0 0

205 8 17 53.00 1 0 1 0 0 0

205 8 23 61.00 2 0 1 0 0 0

205 8 23 59.00 3 0 1 0 0 0

206 8 46 1.184 3 0 1 0 0 0

206 24 32 55.50 50.00 3 0 1 0 0 0

206 24 32 59.50 50.50 3 0 1 0 0 0

206 24 32 53.50 50.50 1 0 1 0 0 0

Doc# Send# Rec# Truth% Acc% d Med Mot Prep Base Inter Exp

Note: Doc# = Document # (as listed in Appendix A), Send# = Number of Senders, Rec# = Number of Receivers, Truth% = Percent Truth Classifications, Acc% = Percent Correct Lie/Truth Classifications, d = Rating Scale Standardized Mean difference, Med = Deception Medium (1=Video, 2=Audio, 3=Audiovisual, 4=Within-Receiver Manipulation, 5=Other), Mot = Sender motivation (0=None, 1=Some, 2=Within-Receiver Manipulation), Prep = Preparation Time (0=None, 1=Some, 2=Within-Receiver Manipulation), Base = Baseline Exposure to Sender (0=No, 1=Yes, 2=Within-Receiver Manipulation), Inter = Interaction (0=Sender is Not Interacting, 1=Sender is Interacting with Receiver, 2=Sender is Interacting with Someone Else), Exp = Expert Receiver (0=No, 1=Yes).

To: Evidence Course Home Page

To: Home Page for Advanced Evidence

Characterizations of Deception

A Double Standard

Research on Detection Accuracy

The Present Review

Determinants of Accuracy

The Double Standard

Light, R.J., Singer, J.D., & Willett, J.B. (1994). The visual presentation and interpretation of meta-analyses. In H. Cooper and L.V. Hedges (Eds.) The handbook of research synthesis (pp. 439-453). New York: Russell Sage.

Number of Senders 1 200 22.45 16.00 22.63

Categorical Variables

Variable # (%) of Receiver Samples Percent Coding Agreement

Deception Medium 91.3%

Other 22 (4.9%)

Sender preparation time 81.1%

Baseline exposure 91.3%

No exposure 360 (93.7%)

Interaction with receiver 33 (8.6%)

Receiver expertise 100%

Table 2

Deception in Three Media: Within and Between Studies

k Raw M 2 (95% CI) Adjusted M

k Raw M (95% CI) Adjusted M

Accuracy d

k Raw M (95% CI) Adjusted M

Table 3

Motivated and Unmotivated Deception Within and Between Studies

k Raw M (95% CI) Adjusted M

No motivation 177 53.36% (± .21) 53.43%

No motivation 231 .462 (± .026) .396

k Raw M (95% CI) Adjusted M

No preparation 177 53.18% (± .21) 53.13%

Preparation 130 53.70% (± .26) 53.75%

No preparation 217 .439 (± .029) .403

k Raw M (95% CI) Adjusted M

No exposure 250 53.35% (± .18) 53.06%

No exposure 331 .400 (± .022) .356

k Raw M (95% CI) Adjusted M

No interaction 85 52.56% (± .27) 52.60%

No interaction 127 .375 (± .036) .302

k Raw M (95% CI) Adjusted M

Non-expert 250 53.31% (± .17) 53.29%

Non-expert 338 .380 (± .022) .387

Mean percentage truth judgments Mean percentage correct lie/truth judgments

Mean Percent Correct by Number of Judgments

0 10000 Appendix A

k Raw M ² (95% CI) Adjusted M

0

10000

Appendix A