Effective Advocacy-Taking It to the Next Level

Effective Advocacy—Taking It to the Next Level

Edward P. Schwartz, Ph.D.

DecisionQuest

800 South St Ste 190 Waltham, MA 02453 (781) 891-8300 [email protected] Edward P. Schwartz, PhD, MSL, is a Consultant in DecisionQuest’s Boston office. In his role, Dr. Schwartz provides quantitative and qualitative analysis of pretrial jury behavior—from interviews and focus group studies to mock trials and large-scale statistical analyses—to provide clients with feedback on case themes, strategies, evidence, witnesses, and presentation style. Further, he consults on case evaluation, provides advice on trial strategies, and assists with jury selection as well as post-verdict juror surveys and interviews. Dr. Schwartz is a nationally recognized jury consultant with excellent analytical acumen and strong market research skills who is noted for his ability to blend the strategic focus of game theory and decision theory with the real-world insights of social psychology to gain a complete picture of how people absorb, analyze, and process information. Dr. Schwartz has conducted jury behavior research, aided with witness preparation, consulted on trial strategy, and assisted with jury selection on dozens of cases, including several high-profile criminal trials. Dr. Schwartz is regularly asked to provide insight on jury trials in the news, having been interviewed by CNN, the Associated Press, the New York Times and countless regional media outlets. Effective Advocacy—Taking It to the Next Level

Table of Contents I. “He Said What?”: Deception Detection and Employment Litigation—Part I: The Gullible Juror...... 5 II. “He Said What?”: Deception Detection and Employment Litigation—Part II: The Fallible Juror...... 14 III. Less Is More? Detecting Lies in Veiled Witnesses...... 21

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 3

I. “He Said What?”: Deception Detection and Employment Litigation—Part I: The Gullible Juror

“He Said What?”: Deception Detection and

Employment Litigation – Part I: The Gullible Juror

ometimes people lie. And yes, sometimes people lie under oath in a court of law. Despite S having taken an oath to “tell the truth and nothing but the truth,” witnesses aren’t always honest and forthcoming – and jurors understand this. The court does too, as evidenced by jury instructions regarding judging the credibility of witness testimony. To wit, consider this language from the Illinois Model Jury Instructions regarding witness credibility:

“You are the only judges of the credibility of the witnesses. You will decide the weight to be given to the testimony of each of them. In evaluating the credibility of a witness, you may consider that witness' ability and opportunity to observe, memory, manner, interest, bias, qualifications, conclusions, the Court anticipated that “If he were allowed to testify, a experience, and any previous the criminal defendant would soon lose guilty defendant might of course inconsistent statement or act by his life. That fact alone didn’t seem to perjure himself in an effort to the witness concerning an issue particularly bother anyone; however, avoid conviction. At the important to the case.”1 the authorities did not want the Founding, many believed that defendant to die having just sinned lying under oath was an especially The temptation to lie in court in order against God by lying under oath. As grievous offense against man and th to protect oneself has always been a such, until the 19 century, criminal God…. Alas, a liar might lose his 2 source of concern and the subject of defendants in the British common law soul even if he saved his skin.” judicial notice. Historically, the courts system were not permitted to testify were very worried about placing a under oath at trial. The prohibition against self-interested criminal defendant’s soul in peril by testimony under oath was not limited asking him to swear to God to tell the As Akhil Amar succinctly states in his to criminal defendants. Until the mid- th truth in court. Since most crimes were recent book, America’s Unwritten 19 century, interested parties to civil capital in the Middle Ages and well into Constitution: The Precedents and Principles litigation were similarly precluded from the Enlightenment, and most We Live By, convictions were foregone

1 A copy of these model instructions can be http://www.state.il.us/court/circuitcourt/civilj 2 From Chapter 3, as excerpted in Amar, Akhil found at uryinstructions/1.00.pdf Reed. “America's Lived Constitution,” 120 Yale Law Journal 1745 (2011).

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 5 testifying under oath.3 As Blackstone expert who honestly believes the conducted in a variety of settings over noted, attorney’s client is in the right than for decades of research, the primary, and an expert who is willing to testify remarkably robust, result is that we, as “All witness, of whatever religion contrary to her own honest opinion. a species, do little better than flipping a or country, that have the use of coin when it comes to detecting their reason, are to be received There are cases, however, in which deception. The average success rate at and examined, except such as are dispositive issues revolve around telling truth from fiction is about 54%.7 infamous or such as are interested in whose version of a key event most The second key finding from the the event of the cause.”4 resonates with the jury. That is, the experimental research is that we all plaintiff says one thing happened and think we are much better at So, while the adversarial system that the defense counters with a different discriminating honesty from lies than characterizes the Common Law story. The jury is asked to resolve a “he we really are.8 This result has profound tradition is touted as an effective said – she said” dispute.5 Employment implications on at least three fronts. method of truth revelation, everyone litigation, especially that involving an First, those in human resources, or who has always understood that interested allegation of discrimination, tends to manage others in the workplace, need parties would sometimes – if not often involve this kind of dispute. On the one to be cognizant of the very real – lie in court. hand, the dissatisfied employee (or possibility that they make regular former employee) claims that a mistakes when evaluating the veracity This is not to say that all cases are supervisor or fellow employee made a of claims made by employees under created equal and that all litigants have particular statement or behaved in a their purview. Second, attorneys equal incentive or opportunity to particular way. On the other, the should not presume that they can effectively misrepresent facts in court. supervisor or human resources correctly ascertain whether a friendly Many types of disputes revolve around manager counters that either the witness is being completely long, ongoing relationships with statement/behavior never took place forthcoming and honest with them in extensive “paper trails.” As such, it is or was completely misinterpreted by preparation for trial. Finally, litigators difficult for a witness to testify to the complaining party. The jury is faced should never rely on jurors’ abilities to certain facts that are simply with the unenviable task of correctly sort out a “he said – she said” inconsistent with too much evidence to determining who is telling the truth – dispute to win a case. There are ways to the contrary. It is hard to argue against and who is lying under oath. increase the likelihood that the jury will accounting statements, signed correctly detect truth and deception, contracts or letters written in one’s own This article (first in a two-part series)6 but the case must have a strong hand. is devoted to reviewing what we know foundation in other areas. about people’s tendencies to lie – both Other kinds of disputes involve more generally and about issues related to the The final empirical result from the interpretive questions than factual workplace – and about our own literature of interest to us here – and ones. Many torts cases require the jury abilities to correctly identify who is the focus of the second article in the to evaluate what constitutes lying to us and who is telling the truth. series -- is that people regularly look to “reasonable care,” even when there is Using these results as a foundation, I exactly the wrong indices of deception. little disagreement among the parties will discuss along the way how issues of We have all heard someone ask that about who did what when. While an deception detection should be handled another person, “look me in the eye expert might lie about his opinion in employment litigation. and say that.” It turns out that liars regarding what qualifies as reasonable typically do a pretty good job of care, this presents the jury with a There are several key lessons here for controlling their facial expressions different type of credibility dilemma human resources professionals and while truth-tellers pay less attention to than the possibility of a litigant lying on attorneys working in the employment the emotional expressions associated this stand. Most experts build their litigation arena. First and foremost, with their speech.9 As such, focus on a reputations and their practices on being human beings are supremely bad at speaker’s face during message reliable, credible witnesses. So, differentiating true statements from transmission actually reduces a listener’s attorneys more often shop for an lies. Over hundreds of studies, ability to tell whether she is being told

3 Ibid at 1746. Maine became the first 5 The genders of the pronouns here are purely 8 See Vrij, A., Granhag, P. A., & Porter, S. American state to permit criminal defendant arbitrary. The point is only that two interested “Pitfalls and Opportunities in Nonverbal and sworn testimony in 1864. Such testimony fact witnesses will testify to different versions Verbal Lie Detection.” 11 Psychological Science in under oath was not permitted in Federal Court of events. the Public Interest, 89–121. (2010) until 1878. Georgia was the final state to 6 The companion piece will appear in an 9 See Vrij, A., Granhag, P. A., & Porter, S. permit the practice, sometime shortly after the upcoming issue of HR Advisor. (2010). turn of the 20th century. 7 Bond, C. F., & DePaulo, B. M. “Accuracy of 4 3 Blackstone, Commentaries on the Laws of deception judgments.” 10 Personality and Social England, 369 (1769). Psychology Review, 214-234. (2006)

6 ■ Women in the Law ■ February 2017 the truth or lied to.10 Consider a typical part of their jobs. This group included witness stand – a low chair behind a tall The most common method of police officers, detectives, judges and screen, so that little more than the determining a baseline accuracy for psychologists. This separation did not witness’s face is visible. This is a terrible human deception detection is to matter for success at deception recipe for accurate deception detection aggregate results across as many detection. All three groups performed by jurors. There are strategies, comparable studies as possible in what about the same, suggesting that the however, that can be used to increase is referred to as a meta-analysis. It is, in professionals were so in name only.13 the likelihood that jurors will recognize short, an analysis of pre-existing the veracity of an honest witness, and analyses. Over the years, Bella A. Trusting Souls: Truth Bias others that can improve the detection DePaulo, of the University of Virginia, in Veracity Evaluation of deceptive testimony. While there is has contributed as much as anyone in some mention of such concerns below, the field to this question. A 2006 One of the major contributors to poor this well-studied and nuanced topic will study11 aggregates results over more performance in these deception primarily be addressed in the second than 200 studies and finds that humans detection studies is the propensity of article. correctly identify the veracity of a humans to believe what they are told. message approximately 54% of the That is, we have a truth bias.14 As I. The Randomness of time.12 That is, if a person chose to mentioned above, these studies Deception Detection completely ignore everything about the typically involved half true statements content of the message and the and half falsehoods. Respondents, Over the years, psychologists and circumstances surrounding its delivery, however, identified statements as true sociologists have conducted hundreds and instead flip a coin to distinguish approximately two-thirds of the time.15 of studies aimed at exploring the ability truths from lies, that person would This is an interesting finding, in and of of humans to distinguish honest perform just about as well as the itself, for what it might tell us about the messages from deceptive ones. It is average participant in any of the human condition. One hypothesis that important at this point to emphasize included studies, all of whom were has been put forth for truth bias is that, that, for our purposes here, a deceptive trying their best to correctly ascertain in our everyday lives, we mostly message is one that the bearer of the whether they were being told the truth encounter people who are honest with message (be it a job applicant, or being lied to. This is a rather bleak us. As such, a person is inclined to significant other, business associate or finding, and one that surprises many believe most statements (without witness) believes to be false, but is people. tangible reason to do otherwise), even delivered in a way intended to convince if primed to be suspicious by the the recipient of the message that it is These meta results are quite robust to revelation that he is participating in a true. A truthful message, by contrast, is variations in how the data are deception detection study. Another believed to be true by the speaker, who aggregated. In the same year, Aamodt hypothesis behind the truth bias is that similarly attempts to convince the and Custer conducted their own meta- we are socialized to be trusting of one recipient that it is true. As such, this analysis of 108 studies, paying special another, through parenting, schooling, review does not cover mistakes, mis- attention to who constituted the religious education and social norms. rememberances, uncertainty on the respondent pool. They separated As such, regardless of whether it is speaker’s part, exaggerations or studies that used college students (very rational to believe that most statements expressions of overconfidence. All of popular source of cheap subjects for are true, an average person does not these other forms of “falsehoods” are, psychology studies) from those that of course, very important, both in used the more general adult population. everyday business dealings and A third category of studies involved litigation; they are however beyond the “professional lie detectors,” defined as scope of this article. people trained to root out deception as

10 Warren, G, Schertler, E and Bull, P., 13 Aamodt, M. G., & Custer, H. “Who can best trust.” In M. L. McLaughlin (Ed.), 9 “Detecting Deception from Emotional and catch a liar? A meta-analysis of individual Communication yearbook, 377-389 (Beverly Hills, Unemotional Cues,” 33 J Nonverbal Behav 59–69 differences in detecting deception.” 15 The CA: Sage, 1986). (2009) Forensic Examiner, 6-11 (2006). The professional 15 See the various meta analyses, all of which 11 Bond, C. F., & DePaulo, B. M. (2006). lie detectors actually out-performed the report similar percentages of truth 12 It should be noted that the standard research laypersons by one percentage point, 55% to identification. Aamodt, M. G., & Custer, H. protocol for deception detection studies is to 54%, a difference that was not statistically (2006); Levine, T. R., Park, H.. S., & expose message recipients to ½ lies and ½ true significant. As a point of reference, the McCornack, S. A. (1999). “Accuracy in statements. For reasons outlined below having Aamodt and Custer study aggregated over detecting truths and lies: Documenting the to do with truth bias, accuracy would vary in 14,000 observations. ‘veracity effect.’” 66 Communication Monographs, predictable ways when the ratio of lies to truths 14 This term was first coined in McCormack, S. 125-144; and Bond, C. F., & DePaulo, B. M. is moved away from this 50-50 split. A., & Parks, M. R. “Deception detection and (2006). relationship development: The other side of

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 7 wish to be perceived as overly cynical the evidence and testimony of have their own set of issues. There is and suspicious.16 coworkers strongly suggests that he usually a history of difficult interactions received an email inviting him to the with coworkers and/or clients. When Truth Bias is a double-edged sword meeting, followed by a reminder two the employer asserts that the when it comes to litigation. On the one days later…” employee’s behavior on the job hand, it is comforting to know that contributed to his negative jurors generally enter the courtroom B. Deny, deny, deny: Lie Bias performance evaluation the plaintiff inclined to believe what they are about exists, as well may be forced to deny this at trial. The to hear, especially when one is resulting denials fall on a skeptical convinced that truth is on one’s side. While truth bias is a very robust result audience. When these tables are turned, On the other hand, if one anticipates in meta-analyses of deception the lie biases of jurors work to the that either counsel or witnesses for the detection, its prevalence does depend advantage of the defense and the other side will attempt to mislead jurors on the type of message being received. disadvantage of the plaintiff. These with false statements, it is somewhat This proves particularly relevant in the results suggest that defense counsel in disconcerting to know how context of witness testimony at trial. such suits should take every undiscerning we humans can be. Vrij and Baxter examined whether opportunity to force the plaintiff to truth bias showed up as often in issue denials on the witness stand. The literature is fairly silent regarding response to denials of alleged behavior what people do when confronted with as it did in response to assertions of C. Truth Bias and the contradictory statements, such that the facts.17 The authors discovered that Incentive to Lie listener must conclude that at least one while subjects exhibited truth bias as party is lying. Given the robustness of expected in response to elaborations of The fact that humans seem to be truth bias, one might imagine that facts and behaviors, they actually socialized to exhibit truth bias means many jurors attempt to interpret exhibited lie bias in responses to denials. that their accuracy of veracity detection seemingly incompatible statements in That is, people tend to assume that depends very heavily on whether they ways so as to maximize the inherent assertions are true and denials are false. are being told the truth or being lied to. truth between them. Examples of this Suppose that a person were completely are often seen in mock trials conducted Unlike the results that we have undiscerning with respect to messages in actual cases. Some mock jurors are reviewed so far, this result would seem he received and chose to believe so averse to calling someone a liar that to have asymmetric implications for everything he heard. He would they twist witness statements quite employment litigation. Typically, the correctly identify every truth that was dramatically so as to be able to believe plaintiff in such litigation is asserting a told to him. He would be considered that both parties are essentially telling positive proposition. “Here is what my statistically to be an excellent truth the truth – or at least trying to do so. A boss did to me.” By contrast, the detector. On the other hand, he would common juror reaction is to interpret a representatives of the defendant never correctly identify deception. He false statement as an error, rather than company are put in the position of would be the world’s worst lie a lie. As a result of this natural issuing denials. “No, that is not what detector.18 tendency of jurors, trial lawyers who happened.” Or, “I never said that.” are overly aggressive in naming a lie or The Vrij and Baxter study would Regular folk aren’t quite as gullible as labeling a witness a liar often suggest that jurors are more likely to our hypothetical example, but we do experience a hostile reaction from believe plaintiffs in this scenario than tend to identify statements as true jurors. A litigator in an employment they should and are more likely to be about two-thirds of the time in studies dispute would do well to put forth a suspicious of company representatives where the actual population of theory of liability that does not rely on than they should. That is, this particular messages only contains one-half true jurors concluding that the other party is combination of truth bias and lie bias statements. The result is that subjects in an out-and-out liar. When possible, works to the systematic advantage of these studies correctly identify about employ language that makes it clear plaintiffs in employment lawsuits, 67% of true statements as being true. that what matters is the falsity of the especially those alleging discriminatory That is, they call an honest speaker a witness’s statement, not the motivation behavior. liar only about one-third of the time. for it being false. E.g.: “Regardless of On the other hand, these same subjects whether Mr. Smith believes that he was Such suits are rarely one-sided, spot a lie only about 41% of the time.19 excluded from the meeting on July 8th, however. Most plaintiffs in such suits This means that a liar successfully

16 Levine, T. R., Park, H.. S., & McCornack, S. and Individual Differences,” 7 Expert Evidence believer discussed here would reveal no false A. (1999) at 129. 25-36 (1999). positives but would be very prone to false 17 Vrij, A., and M. Baxter, “Accuracy and 18 For those of you who are familiar with negatives. Confidence in Detecting Truths and Lies in probability theory, these can be thought of as 19 Levine, T. R., Park, H.. S., & McCornack, S. Elaborations and Denials: Truth Bias, Lie Bias Type I and Type II errors. The hypothetical A. (1999).

8 ■ Women in the Law ■ February 2017 convinces his audience that he is telling “How could I not have seen this In our social and business worlds, we the truth well over half the time. We coming?” is the common lament. One communicate with each other in many would all do substantially better at lesson from the extensive study of different forms. Face-to-face speech detecting lies if we ignored what we saw deception detection is that most people has been around for the longest, but and heard and flipped coins instead!20 are inexpert at knowing when they are many methods are available to being lied to. As such, a committed liar communicate over distances, from It is not uncommon for an attorney to would surely have succeeded just as written and types letters, to telephone, express the conviction that a key well under someone else’s watch. to email and text message. Many witness for the other side is a big fat scholars have examined both our liar. The attorney seems incredulous Another thing to keep in mind is that propensities to lie via various media, as that anyone buys what this person is lying during job-hunting appears to be well as our ability to discern true selling. Many attorneys seem confident rampant and widely accepted. statements from false ones depending that jurors will see right through the According to recent studies, up to on how a message is delivered.22 liar’s deception – how could they not? three-quarters of job applicants Well, the cautionary tale from the admitted to engaging in deceptive In some ways, immediacy has much to hundreds of studies conducted so far communications (including false recommend it, in terms of the ability to about deception detection is that many, statements on resumes, exaggerating deftly deliver a lie. Face-to-face many jurors will not be able to tell grades and salaries during interviews, communication permits the speaker to when they are being lied to. Keep in and identifying competing offers that add nuance and inflection in order to mind, however, that most of these did not exist).21 According to the Prater be maximally convincing. On the other studies involve statements made by and Kiser article, roughly 90% of hand, such intimate communication strangers to strangers, with no context undergraduate job applicants admitted forces the speaker to confront the or other cues about the veracity of the to using some form of deception human consequences of his deception. statements being made. One job of the during job-hunting. In much the same way that soldiers find trial team is to provide the context and it easier to shoot each other from a supporting materials to facilitate the In light of these two factors, managers distance, a liar is likely to suffer lower ability of jurors to distinguish fact from and human resource professionals psychic cost from lying at a distance. fiction. The lesson here should not be should not beat themselves up for By contrast, one can lie via a letter or that jurors can’t tell who is lying to failing to identify a particular dishonest email and perhaps never be forced to them, but rather that they can’t be employee. The odds are frankly stacked confront any of the harm suffered by relied upon to get there without help. against them, given the prevalence of the recipient. In some circumstances, it One goal of these articles is to provide lying during the job-hunting process, might be possible to practice deception guidance about how to give jurors the the limited resources available to verify anonymously. Perhaps people lie most help they need to believe the truth- submitted information and the often in email or letters. The problem tellers and doubt the liars. inherent difficulty humans have with practicing deceit via the written knowing when they are being lied to. word is that the recipient is exposed to When an employment relationship The real challenge appears not to be so no indexes of sincerity. The written lie sours, it creates an unpleasant much to weed out the liars (for no must prevail exclusively on the quality circumstance for employer and company would seem capable of of its content. A statement that rings employee alike. It is only natural to functioning without them) but to hollow on paper cannot be salvaged by speculate on what could have been identify those whose lies are of a nature tone or emotion. Lying is writing also handled differently along the way to and frequency to suggest that they brings with it special dangers. It is very avoid the situation that ultimately led to cannot be trusted to do their jobs hard to back away from something litigation. It is not uncommon for a consistent with the company’s needs recorded in writing or explain it away human resource professional or and trust. later as a misunderstanding. As such, longtime manager to express dismay at while lying via email or letter might be having trusted a colleague or employee D. Some Lies are All in the tempting in terms of its low psychic who turned out later to be unreliable, Telling cost, it is very risky in terms of being unscrupulous or simply dishonest. exposed as a liar.

20 Albeit, only in a world where we got lied to Kiser, S. B. “Lies, lies, and more lies.” 67 SAM (New York: Association for Computing half the time but did not know in advance that Advanced Management Journal, 9–36 (2002). Machinery, 2004), as well as, DePaulo, B. M., the proportion of lies was set at that level. 22 See generally, Hancock, J., Thom-Santelli, J., Kashy, D. A., Kirkendol, S. E., Wyer, M. M., & 21 Levashina, J., & Campion, M. “Measuring & Ritchie, T. “Deception and design: The Epstein, J. A. “Lying in everyday life,” 70 faking in the employment interview: impact of communication technology on lying Journal of Personality and Social Psychology, 979-995 Development and validation of an interview behavior,” In E. Dykstra-Erickson, & M. (1996). faking behavior scale.” 92 Journal of Applied Tscheligi (Eds.), Proceedings of the 2004 conference Psychology, 1638–1656 (2007). Prater, T., & on human factors in computing systems, 129-134.

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 9 the speaker. We do much better at else to look at during the witness’s Empirical studies have actually ignoring deceptive verbal cues than testimony. Rather than just handing a discovered that we lie most often by deceptive non-verbal ones.26 document to the witness when asking if telephone. 23 This method of deception he has ever seen it, put a copy up on a seems to be more a compromise than We are left then with a rather strange big screen in front of the jury. an ideal choice. A telephone call would set of results. People lie most often Encourage the jury to listen only with seem to be neither the most effective during voice-only communications, their ears. We will spend some time on nor the safest way to tell a lie. It does, like phone calls, but doing so this “distraction theory” in the next however, mitigate fairly well the maximizes the likelihood of being article. shortcomings of other options. One (accurately) disbelieved. It is important can control the tone and tenor of the to keep in mind, however, that the E. Boys will be Boys … message through ones voice (but not initial convincingness of a message is Unfortunately through one’s expressions) and doing only one aspect to a successful lie. The so does not leave a “paper trail.” liar needs to be able to sustain the ruse Men and women approach the and also avoid getting trapped in the lie prospect of lying – and being lied to – Somewhat ironically, empirical studies due to subsequent events or quite differently. Li Li wrote a doctoral of deception detection generally show statements. With these considerations thesis at the University of Miami that humans are relatively good at in mind, it might still be rational for a dedicated to exploring gender distinguishing truth from fiction when would-be liar to employ the telephone differences in lying and deception listening to vocal messages.24 The as his instrument of deception. detection.27 Among the studies cited by reason for this is that people do a much Li in an extensive literature review is a better job of employing verbal cues of We will discuss at length the 1992 exploration of how men and deception than non-verbal cues of implications of the “cues to deception” women generally feel about lying.28 deception. While this intrudes a bit on literature in the next article, but I would According to Li, “Their study indicates the subject of the next article, it is like to point out here just a couple of that regardless of the lies’ content, worth taking a moment to explain a bit strategies to consider. When a manager targets, and relationship with the liars, about such cues. When a person or HR professional is reviewing his women rate deception as more attempts to evaluate the veracity of a notes about an employee as part of significant, more unacceptable, and statement, he naturally evaluates many preparation for trial, it would be reported significantly more negative aspects of the message he is receiving. worthwhile to focus on telephone emotional reactions toward discovered Does the content make sense? Are the conversations with the employee that lies.” This visceral aversion to lying is word choices strange? Is there struck the manager as odd, important to keep in mind when hesitation or quavering in the voice? incongruous or misleading. Given our considering how women differ from Does the speaker make eye-contact? relative prowess at detecting deceit men in terms of deception detection. Does he fidget? Do his facial through verbal cues, digging deeper expressions and tone of voice match into the subjects of those conversations Relatively few studies paid much what he is saying? Some of these cues could reveal useful evidence. The attention to gender difference in are verbal and others are not. The second strategy involves getting jurors deception detection prior to Li’s thesis. studies are consistent in finding that we to focus on verbal cues when a witness One study that looked at deceptive humans are easily duped by the non- is testifying. Especially if one is statements about art appreciation (so verbal cues on which we tend to focus concerned that a witness is a “good that there was no gender-based most.25 Absent access to these non- liar,” in that he controls his facial content) found that women were verbal cues, we are forced to actually expressions well, it might prove significantly more likely than men to listen to the message being delivered by beneficial to give the jury something believe deceitful statements.29 One

23 Where “most often” is defined by a & O'Brien, T. P. (1988). “The motivational 27 Li, Li, “Sex Differences in Deception percentage of any given kind of message that impairment effect in the communication of Detection,” Open Access Thesis, Paper 261. contains intentional mistruths. Studies of lie deception: Replications and extensions.” 12 (2011) frequency are typically conducted by having Journal of Nonverbal Behavior, 177-202 (1988). 28 Levine, T. R., McCornack, S. A., & Avery, P. subjects keep very detailed diaries of their daily 25 See Vrij, A., Granhag, P. A., & Porter, S. B. “Sex differences in emotional reactions to communications and indicate, in each instance, (2010) for an excellent treatment of verbal and discovered deception.” 40 Communication whether they told a lie. While the diaries are non-verbal cues to deception. Quarterly 289-296 (1992). subsequently reviewed and coded 26 As will be discussed in the companion 29 DePaulo, B. M., Epstein, J. A., & Wyer, M. anonymously, such studies are fraught with article, not all non-verbal cues are created M. “Sex differences in lying: How women and data reliability concerns due to self-reporting equal. Some forms of body language, as well as men deal with the dilemma of deceit,” In M. by subjects. See both Hancock, J., Thom- emotional “leakage” in facial expressions can Lewis, & C. Saarni (Eds.), Lying and deception in Santelli, J., & Ritchie, T. (2004) and DePaulo, be useful tracers of deception, provided everyday life, 126-147. (New York: Guilford et al. (1996). listeners can be trained to focus on them. Press, 1993). 24 For an excellent treatment of this subject, see DePaulo, B. M., Kirkendol, S. E., Tang, J.,

10 ■ Women in the Law ■ February 2017 economics experiment involved the The men, on the other hand, seemed to skeptical and will particularly fail to division of a fixed monetary reward.30 confound their audience, with fewer recognize honesty from another man One subject would tell his or her than 40% of respondents correctly on the stand. Men seem to do a slightly partner which portion (A or B) distinguishing truth from fiction. Why better job of recognizing dishonesty contained the larger share. The partner were the women so much more among female speakers, but nobody would then decide whether or not to transparent as messengers? seems particularly well-equipped to believe the message (and pick a portion spot a male witness who is lying to accordingly). While subjects from both The short answer is that people them. We will return to this question genders were equally good at generally believe women are honest later, after we have discussed cues to distinguishing true messages from lies, and men are liars. Female subjects deception and the role context in the male subjects lied to their partners correctly identified truthful women evaluating message veracity. substantially more often (55% versus speakers over 83% of the time. Men 38%). Consistent with other studies, even picked up on honest women over II. Unwelcome Bravado: there was a substantial truth bias, in 75% of the time. So, as a general rule, Overconfidence in deception that men and women believed their honest women are transparently so. detection partners well over 75% of the time Both women and men correctly despite being lied to almost half the identified lying women just about 55% In jury deliberations one often time. of the time (with men getting it right encounters the phenomenon of the slightly more often). Honest men had “false expert,” the person who is Motivated by these intriguing results, Li great difficulty conveying their confident in his own knowledge of a sought to explore whether men and sincerity, convincing subjects to believe subject or prowess at some analytical women performed differently as them slightly more than half the time task, but whose actual knowledge base deception detectors, especially with (men actually believed honest men or talent is just as suspect as everyone respect to the gender of the messenger. slightly less than half the time). The else’s, if not more so.31 False experts are The results of Li’s study are quite truly astonishing result is that lying men a particular problem for jury dramatic. As with earlier studies, Li succeeded in fooling their audience deliberations for a number of reasons. confronted subjects with half true three-quarters of the time. To put that First, the subject matter is usually quite messages and half lies. Half of the in perspective, a subject was almost twice as foreign to the participants, as is the messages were delivered by men and likely to identify an honest man as a liar as he decision task. As such, jurors are half by women. The overall accuracy or she was to identify a deceitful man as a liar. generally relieved at the prospect of rate was 54%, just in line with previous Whatever cues of deception these someone in the room having some clue studies. subjects are looking to, they are clearly as to what they are doing. Second, since the wrong ones and their observations the jurors are all strangers to each Women showed greater truth bias, are leading them in exactly the wrong other, each cannot rely on past identifying 63% of statements as true, direction. In the next section of this experience to evaluate the likelihood while their male counterparts only article, we will explore what empirical that any professed expert in their midst thought that 59% of statements were research can teach us about cues for actually has any expertise. Finally, the true. Almost all of the added truth bias honesty and deception. jurors are limited to considering as for the female subjects came in the evidence only what has been presented form of correctly identifying truthful Before moving on, this would be a to them at trial. They are not permitted statements. That is, both groups good place to reflect on how the results to conduct any independent research to spotted liars about 41% of the time, but discussed so far might be relevant for determine whether the opinions women correctly identified truth-tellers jury selection in employment cases. professed by the “expert” have any at a 70% clip, as opposed to 60% for The discovered gender differences merit. This is a dangerous recipe for a their more suspicious male suggest that truth identification by verdict to be “hijacked” by someone counterparts. jurors can depend very much on the who is overconfident in his knowledge gender of the witness under scrutiny, as and evaluation of issues relevant to the The most interesting gender effect well as whether that witness is likely to case at hand. involves the identity of the speaker. be telling the truth on the stand. While Women were substantially easier to all jurors are likely to detect the One of the most common subjects read, with respondents correctly sincerity of a truthful female witness, about which we see false expertise is identifying the veracity of their women would seem to do so more how the human resources process messages almost two-thirds of the time. easily than men. Male jurors will be works at various kinds of companies.

30 Dreber, A., & Johannesson, M. “Gender 31 Such false experts are not limited to the jury groups, PTA meetings and around the kitchen differences in deception,” 99 Economics Letters, room, of course. We have all encountered table. 197-199 (2008). them in board rooms, class rooms, study

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 11 Jurors claim to understand what the more circumspect fellow jurors, the I leave you here with three important law requires an HR professional to do overconfident ones are likely to rely factors to consider in anticipation of in various situations; usually, that more heavily on such an assessment the companion article to this one. First, person is wrong. Jurors express than they should. If a juror decides cases do not typically revolve unwarranted confidence that a early on that a witness is a liar, he might exclusively around whether the jury company would have a particular reach a conclusion about the correct believes one side of a “he said – she human resources policy in place, verdict in the case then and there, said” scenario or the other. There is notwithstanding the absence of any foreclosing his mind to reliable typically presented lots of evidence evidence regarding such a policy. Jurors evidence. An overconfident veracity about work history, performance also estimate with unjustified certainty detector will weigh competing evaluations, prior complaints by the what a compensation or severance or testimony inappropriately and will be plaintiff or others, financial statements pension package would look like for a resistant to arguments contrary to his and more. The disputed statements or given employee. All of these kinds of initial assessment. As such, during voir events comprise just one piece of a false expertise can pervert jury verdicts dire, one would be wise to challenge much bigger puzzle to be solved by the in employment cases. prospective jurors who believe jury. In general, juries get most of the themselves to be very skilled at other stuff “right.” But what of confidence in deception detecting deception. Such jurors detection? Are some people just better introduce variance into the process Second, a disputed statement is not at it than others? The answer is, “Yes, without any increase in accuracy. made in a vacuum. Unlike in most of some people are, in fact, better than the studies on deception detection, others at discerning truth from lies.” III. All Hope is not Lost: when a juror is confronted with a The problem is that there is essentially Strategies for improving juror statement by a witness, that juror no relationship at all between a performance as deception detectors generally is exposed to information person’s internal belief about his ability about the situation, the facts underlying to detect deception and his actual The lessons from the research on the questionable testimony and the ability to do so. As I mentioned above, detection deception enumerated above character of the witness. That is, the “professional” lie detectors do not fare would seem to paint a rather bleak statement can be placed in a context any better than laypersons at telling picture of the ability of jurors to know that can be used to help evaluate its truth from lies, notwithstanding their who is telling the truth and who is lying veracity. Much of the evidence and sometimes extensive training in doing on the stand. It is not encouraging to testimony presented at trial is an effort so.32 These professionals, on the other know that the average person performs by the parties to put the underlying hand, are quite confident that they are little better than a coin-flip in dispute into context. To quote the title superior deception detectors than their distinguishing truth from fiction. of a recently popular book in social lay counterparts.33 Perhaps even more disturbing, we are a psychology, “situations matter.”36 gullible people – motivated liars can Researchers have begun to explore the The false expert problem is not limited fool us six times out of ten.35 Does this important role that context plays in to professionals. In a meta-analysis of mean that an employment case that deception detection and how that 18 different studies that compared revolves around competing versions of context can be shaped to improve confidence in deception detection with events is really just a coin flip? A spin performance. This will be a major topic actual performance, the correlation of the roulette wheel? A throw of the of discussion for the next article. found was just 0.04, which proved not dice? to be significantly different from zero.34 Finally, scholars have been examining That is, people who think they are good The answer, of course, is “no,” for the cues we focus on when trying to at telling truth from lies generally there is much missing from the story, discern if someone is telling us the aren’t. as we move from hypothetical truth or lying to us. As mentioned statements used in academic studies to above, we generally turn to all the In fact, people who think they are good the nuts and bolts of an employment wrong cues, thwarting our own efforts at deception detection are dangerous dispute. to learn the truth; but, it doesn’t have jurors. While they might not mistake a to be that way. There are cues we can truth for a lie any more often than their look for that do signal some kind of

32 See Vrij, A., Granhag, P. A., & Porter, S. Human Behavior, 211–227 (2005). It is also accuracy–confidence correlation in the (2010) for an excellent treatment of verbal and worth noting that there are several studies in detection of deception.” 1 Personality and Social non-verbal cues to deception. which training in deception detection actually Psychology Review, 346–357 (1997). 33 See Kassin, S.M., Meissner, C.A., & degrades performance in distinguishing truths 35 Even more often if they are men. Norwick, R.J. “’‘I’d know a false confession if I from lies. 36 Sommers, Samuel. Situations Matter: saw one’’: A comparative study of college 34 DePaulo, B.M., Charlton, K., Cooper, H., Understanding how context transforms your world. students and police investigators.” 29 Law and Lindsay, J. L., & Muhlenbruck, L. “The (Riverhead Press, 2011).

12 ■ Women in the Law ■ February 2017 cognitive dissonance – a disconnect between the speaker’s words and his beliefs. Such a disconnect provides us with a red flag that something is not copacetic about what we are hearing. It might just be a lie. In the next article, I will review the literature on effective use of verbal and non-verbal cues and discuss how these results can be leveraged into effective trial strategies in employment litigation.

Edward P. Schwartz is a Consultant in DecisionQuest’s Boston office. Dr. Schwartz provides quantitative and qualitative analysis of pretrial jury behavior – from interviews and focus group studies to mock trials and large- scale statistical analyses – to provide clients with feedback on case themes, strategies, evidence, witnesses, and presentation style. Further, he consults on case evaluation, provides advice on trial strategies, and assists with jury selection as well as post-verdict juror surveys and interviews. Dr. Schwartz is a nationally recognized jury consultant with excellent analytical acumen and strong market research skills who is noted for his ability to blend the strategic focus of game theory and decision theory with the real-world insights of social psychology to gain a complete picture of how people absorb, analyze, and process information.

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 13 II. “He Said What?”: Deception Detection and Employment Litigation—Part II: The Fallible Juror

“He Said What?”: Deception Detection and Employment Litigation – Part II: The Fallible Juror

ow often have you heard the phrase, “Look me in the eye and say that.”? It is deeply H rooted in our collective psychology that the best way to tell whether someone is telling the truth is to look them in the eyes while they are speaking. Suppose I were to tell you that you'd be better off closing your eyes while that person was talking to you? Well, that advice is among surprising lessons learned from research on deception detection, conducted by psychologists, sociologists and neuroscientists over the past few decades.

Here are a number of The major reason that they serve so blame her if she fidgets, or stammers, commonly relied upon indicators of poorly as proxies for deception is that or perspires or occasionally loses deceptive behavior: speakers who want to be believed focus? Of course not; but jurors are regularly exhibit such behaviors even likely to interpret these symptoms of  Avoidance of eye contact when they are telling the truth, stress as red flags for deceitfulness.  Stuttering especially in high stress situations.  Perspiration This article (second in a two-part  Fidgeting Consider the HR professional who series)2 is devoted to reviewing what we  Hesitation takes the stand in an employment know about strategies people use to  Blinking discrimination lawsuit. Her own evaluate whether a speaker is being handling of the underlying workplace truthful or deceptive, the actual  Shifting position situation might be in question. She is effectiveness of those strategies, and  Nervous laughter. often speaking as the company's what has been learned over the years representative, a role with a lot of about how to help people improve Whether a person can articulate responsibility and pressure. Her their performance in terms of exactly what he is looking for when testimony might be central to the jury's deception detection. This discussion trying to determine whether he is being evaluation of the case—or she might follows upon what was covered in the lied to, he likely makes note of these just believe that it is. She is an first article and so a brief review is in various visual and auditory cues in unfamiliar environment, being order. conducting his evaluation. The questioned by a hostile stranger, while problem is that we, as human beings, a stern judge in a black robe stares at First and foremost, human beings tend to focus on the wrong things and her from on high. In short, this is an are supremely bad at differentiating also make the wrong inferences from extremely stressful environment. The true statements from lies. Over what we observe. The “tells” on the list witness is highly motivated to be hundreds of studies, conducted in a above have been shown experimentally believed by the jury. Would anyone variety of settings over decades of to be very poor indicators of lying.1

14 ■ Women in the Law ■ February 2017 research, the primary, and remarkably notwithstanding the fact that those I. TRANSPARENCY AND robust, result is that we, as a species, do evaluations are likely to be faulty. A key TRUTH DETECTION little better than flipping a coin when it corollary to this result is that there is comes to detecting deception. The virtually no statistical relationship Over the years, psychologists and average success rate at telling truth between a person's confidence in his sociologists have conducted hundreds from fiction is about 54%.3 ability to tell whether someone is lying of studies aimed at exploring the ability and his actual ability to do so. The juror of humans to distinguish honest In the first article, I reviewed who proclaims in the jury room that a messages from deceptive ones. It is several variants of this basic result. particular witness was clearly lying is no important at this point to emphasize “Professional” lie detectors, such as more likely to have correctly assessed that, for our purposes here, a deceptive law enforcement agents, judges and the witness's veracity than anyone else message is one that the bearer of the mental health professionals, do not on the jury. This result has profound message (be it a job applicant, perform significantly better than the implications for litigators in the significant other, business associate or rest of us at discerning truth from courtroom, in that it is a very witness) believes to be false, but is fiction. We are all generally more dangerous strategy to rely on a jury's delivered in a way intended to convince suspicious of messages from men, and ability to identify which witnesses are the recipient of the message that it is yet they are more effective than women being truthful in order to win a case. true. A truthful message, by contrast, is at duping us with their lies. From a believed to be true by the speaker, who gender perspective, the most accurate People are terrible at figuring out similarly attempts to convince the kind of evaluation is women correctly when they are being told the truth and recipient that it is true. As such, this discerning when another woman is when they are being told lies. These review does not cover mistakes, being truthful. This result needs to be same people are convinced that they misremembrances, uncertainty on the interpreted with a grain of salt, are actually quite competent at speaker's part, exaggerations or however, due to something known as discerning truth from lies; and there is expressions of overconfidence. All of the “truth bias.” Humans, in general, no reason for the self-professed lie these other forms of “falsehoods” are, and women to some extent more than detectors to exhibit such confidence. of course, very important, both in men, tend to believe they are being told This begs the question, “What are we everyday business dealings and the truth more often than they are. In doing wrong?” The answer, and the litigation; they are, however, beyond studies where participants are exposed focus of this article, is that people the scope of this article. to equal ratios of lies and truths, they regularly look to exactly the wrong still identify almost two-thirds of indices of deception, such as those Each message importantly statements as true. As such, just by enumerated above. involves two people: the sender and the sampling, participants will correctly receiver. The well-known philosophical identify a higher percentage of truths With respect to the “look me in the question asks if a tree falls in the forest, than lies. eye” test, it turns out that liars typically and no one is around to hear it, does it 7 do a pretty good job of controlling their make a sound? The philosophical An interesting thing can happen, facial expressions while truth-tellers question can be restated, for our however, when the statements being pay less attention to the emotional purposes, in terms of whether a evaluated are “denials” instead of expressions associated with their sound—or a message—that does not “assertions.” Researchers have speech.5 As such, focus on a speaker's reach any listener can still have detected a “lie bias” with respect to face during message transmission meaning. A great majority of the early denials, which calls into question our actually reduces a listener's ability to tell studies of deception detection focused collective commitment to the idea of whether she is being told the truth or on the ability of the listener to discern “innocent until proven guilty.” Forcing lied to.6 Consider a typical witness truth from fiction. While there was someone to deny something on the stand—a low chair behind a tall screen, some focus on the nature of the stand—even truthfully—can cause so that little more than the witness's message—its subject matter , length, jurors to increase their internal face is visible. This is a terrible recipe syntax, etc.—and how such things estimates of the likelihood that the for accurate deception detection by could affect deception detection person is a liar. jurors. There are strategies, however, performance, relatively little attention that can be used to increase the was paid to the role played by the The second key finding from the likelihood that jurors will recognize the speaker. experimental research is that we all veracity of an honest witness, and A. Demeanor vs. Transparency think we are much better at others that can improve the detection discriminating honesty from lies than of deceptive testimony. These Relatively recently, researchers we really are.4 Therefore, jurors will strategies comprise the bulk of the began to ask whether it might be the often rely heavily on their evaluations discussion below. case that certain speakers are just easier of a witness's credibility, to read than others.8 Might it be the

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 15 ability of the sender to convey effectively they can lie? Or, how on a 7-point scale, ranging from “not at trustworthiness, rather than any innate convincingly they can report the truth? all obvious” (1) to “very obvious” (7). ability of the listener, that contributed Participants rated their own lies as a 3.0 to effective deception detection? This Recent research has revealed that on the obviousness scale, on average, line of inquiry required further we are not appreciably better at while the average response from an elaboration of what could be meant by recognizing our own transparency than observer partner was only 2.0. terms like “credible” and recognizing it in others. Gilovich, “trustworthy.” Surely, a speaker who Savitsky and Medvec (1998) conducted These results have important always came across as honest, even a series of experiments aimed at implications for litigators concerned when she was lying, would not improve exploring whether people have a about witnesses lying on the stand listener accuracy. Listeners would realistic sense of their own during trial. While the general inability correctly identify her true statements transparency.10 In one experiment, of jurors to correctly discern who is but would always be fooled by her lies. students were placed in groups of 5 and being honest and who is lying to them By the same token, a speaker who all asked to answer the same question is extremely troubling, the concern seems to everyone to be “the lying about their own personal experiences. would be somewhat mitigated by a sort” would have his deceptions In each round, one student was belief that witnesses generally tell the detected (a good outcome), but would assigned to be the liar. After all five truth under oath. In addition to deeply have a great deal of difficulty getting students had spoken their answers, the ingrained social norms against lying anyone to believe his honest statements four truth-tellers were asked to identify and a strong commitment to the (a bad outcome). the liar in the group and the liar was integrity of the judicial system, we can asked to estimate the percentage of add another reason to suspect that It became necessary, then, to other group members who would witnesses will be loath to lie on the distinguish between “demeanor” and correctly identify him (or her) as the stand: they think they'll get caught. The “transparency.” Demeanor describes liar. By pure chance, the accuracy rate illusion of transparency can act as a the level of trustworthiness a speaker should have been 25% and the deterrent (albeit a false one) to perjury. conveys, regardless of the truth of his experimental result was 25.6%. message. So, some people just come Therefore, we can conclude that, at C. In Search of Transparent across as honest; others seem shifty. least under these conditions, the Testimony This is a description of demeanor. By student subjects were not very contrast, a person is transparent if he transparent at all. By contrast, the mean There might only be a small seems trustworthy when he tells the estimate of transparency (percentage of portion of the population who are fairly truth and deceptive when he lies.9 my group members would identify me transparent throughout their lives, in Transparency can be thought of as a as the liar) was 48.8%. That is, subjects that their lies are easily distinguishable characteristic of a person, in that he is thought they were twice as transparent from their truths across a wide range of generally easy to read, or of a person in as they were. circumstances. That is not to say, a particular context. For instance, a however, that most of us cannot be person might be very transparent in his The researchers were concerned transparent under some circumstances. communications with his spouse but that the overestimate of transparency The question facing a trial attorney is nontransparent at the poker table. might just stem from the general how to increase transparency for perception that lies are easier to spot witnesses on the stand. Is there some B. Mr. Cellophane—It's all in than they really are. The team then ran method of questioning that will your mind a replication of the study with an increase the ability of jurors to identify important twist. Each subject was both truthful witnesses as credible and From the perspective of the yoked to an “observer.” The observer dishonest ones as liars? listener, we know that people are not knew whether his partner was a truth- very good at all at reading the cues to teller or liar for each round but knew Levine, Shaw and Shulman sought 11 deception. Most people look to nothing more about any of the other to explore this very question. They unhelpful cues and ultimately perform players. Once again, participants hypothesized that asking subjects little better than chance when overestimated their transparency directly whether they were lying could attempting to differentiate lies from (44.3% estimate of being pegged as the lead to greater transparency than truths. What's more, we have liar); however, the observers did not simply asking them the questions to completely unrealistic expectations overestimate their partners' which they might offer lies. It was not about our own abilities to perform this transparency (25.3% estimate of being obvious that such an approach would task. What then of the speakers who caught lying). As a further check on this increase transparency. Asking a liar are sending the messages? Do they result, the researchers had each about his lying puts him on notice of typically appreciate how convincing participant and his observer rate the his veracity being scrutinized, thereby they are? Do witnesses know how “obviousness” of the participant's lie increasing his incentive to mask his insincerity. By contrast, the honest

16 ■ Women in the Law ■ February 2017 person could have a whole host of discovered in other studies—slightly These results bring with them two emotional responses to being accused better than chance. There was the very important lessons for litigators. of lying, some of which might be expected truth bias, with listeners The first is that witnesses vary a great misinterpreted by listeners as some identifying 60% of interviewees as deal with respect to their inherent indication of dishonesty. honest (despite only half of them being transparency. The best way to ascertain so). The key result, however, was the whether a witness is transparent is to The researchers had subjects difference in success rate depending on subject a simulation of their testimony participate in a trivia game with a interrogation method. The subjects (or a clip from a deposition) to a focus partner, ostensibly to explore issues of who only saw background interviews group for reaction. In addition to teamwork. The partner was a only correctly distinguished cheaters asking for general ratings of credibility confederate who would encourage the from non-cheaters 44% of the time. and honesty, it would be wise to ask subject to cheat on the trivia game. On the other hand, the ones who respondents to evaluate the Following the game, each subject was watched the direct and strategic truthfulness of specific key statements interviewed briefly with three varieties interviews had a 68% accuracy rate. related to the central themes of the of questions. Only two types of Cross-examination very substantially case. When it really counts, will jurors questions are relevant for our purposes increased witness transparency. believe your witness? For an adverse here. Each subject was first asked witness, will jurors detect false rather innocuous questions about The results further supported the statements? Will they dismiss even true experience with trivia games and team hypothesis that transparency of the ones? Assuming that your trial team exercises. At the end of the interview, message sender is a greater knows which statements are true and each subject was asked pointedly contributing factor to successful which are false, it will be possible to whether anyone had cheated at the deception detection than is the judging differentiate between a witness's game. Each person who denied skill of the listener. For those judges demeanor and her transparency. One cheating was then asked if he was who watched the background aspect of this is worthy of mention. Just telling the truth. Finally, he was asked interviews, 85% of them had a success as we are very poor evaluators of what his partner from the game would rate between 30% and 50%. That is, veracity, but think we are much better say about whether any cheating had almost everyone was pretty mediocre. than we are, I would strongly counsel gone on. All cheaters who confessed The group who watched the direct attorneys not to assume that they can were excluded from the sample, leaving interviews performed better but had easily discern themselves whether a 22 subjects who cheated and then similarly homogeneous success, with witness will be transparent or not. completely denied doing so during the 85% of them correctly distinguishing Don't take a chance that your coin �ip subsequent interview. cheaters from non-cheaters between turned out wrong—go test out reaction 60% and 80% of the time. Some of the to any key witness to confirm that The researchers matched these speakers, however, proved to be jurors will react to testimony as you lying cheaters with 22 demographically extremely transparent while others expect they will. similar subjects who played the game were completely confounding. 9% of honestly and subsequently denied the speakers were incorrectly identified The second important lesson is cheating (Happily, there were no false as cheaters or non-cheaters more than that direct and strategic interrogation confessions). These 44 interviews were 80% of the time when only their can increase witness transparency. then edited into two distinct sets. The background interviews were observed. Note that this was true even for first set included only the background On the other hand, among those whose subjects who did not cheat and were questions about trivia and teamwork. background interviews were watched, honestly denying the accusation of The second set included only the direct 9% were so transparent that they were having done so. Therefore, while a and strategic questioning about correctly classified more than 80% of tough cross-examination will increase cheating and lying. The researchers the time. This variation in transparency the likelihood that jurors will detect then showed the 44 background carried over to the direct/ strategic deceit from an adverse witness (no real interviews to 64 different subjects, interviews, as well. One quarter of the surprise), asking your own witness to asking each one to identify each speakers were misclassified more than confirm that she is telling the truth can interviewee as either a cheater-liar or an half the time even when facing direct actually increase the likelihood that honest non-cheater. They showed the interrogation. On the other hand, more jurors correctly identify her as being 44 direct and strategic interviews to a than a third of the speakers were honest. There is an important caveat different group of 64 subjects and correctly identified as cheaters or non- here. Because transparency varies asked them to also identify each cheaters by more than 80% of listeners across speakers, direct and strategic interviewee as either a cheater-liar or after watching their direct questioning will not uniformly improve honest non-cheater. interrogations. juror judgments of veracity. For some witnesses, it will be very effective, while Overall accuracy was 55.9%, very much in keeping with what has been

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 17 for others, it might not make much of to hear the speakers words definitely Ekman and some of his colleagues a difference. improved performance in lie detection have developed training techniques to over watching silent footage of the teach people to detect micro- II. IMPROVING LISTENER speaker. The ability to see the speaker's expressions during interrogations, PERFORMANCE body was also somewhat helpful. By interviews and testimony.16 There is contrast, being able to watch the great variation in the effectiveness of Given the very poor performance speaker's face during the delivery of the this training, from person to person, of experimental subjects in deception message actually reduced the ability of but it can improve micro-expression detection studies, researchers have the listener to discern whether he was detection accuracy quite profoundly for turned their attention to identifying being lied to. The speaker's face proved some people.17 what it is, if anything, that observers do to be an unhelpful distraction. wrong that causes them to so poorly Listeners have a tendency to focus on While these results are theoretically distinguish truth from fiction. If we the one part of the speaker that liars fascinating, they do not provide much understand better the errors we make, control better than truth-tellers. to be used in litigation from a practical it might be possible to develop standpoint. Litigators are not permitted strategies for improving accuracy in There are two straightforward to subject jurors to micro-expression distinguishing truths from lies. strategies to overcome this failure in detection training, even were the lawyer the way people try to distinguish truth or his client willing to pay for it. The A. The Wrong Cues from fiction. The first is to get listeners Court will not allow counsel to record As noted above, many common to focus on more productive cues. The testimony and play it back in court in beliefs about how liars behave second is to disrupt the ability of liars super-slow mot ion so that each micro- differently from truth tellers turn out to to control their facial reactions while expression can be pointed out to the be counterproductive. That is, they are testifying. jury. In addition, the researchers who misconceptions. Many of these have pioneered this work are very B. Leakage: Break the Dam misconceptions are related to facial careful to point out that emotional expressions. leakage is not a direct indicator of lying. Paul Ekman and his collaborators Such leakage indicates only that there is pioneered an approach to deception a disconnect between the speaker's true A liar will not look the listener in detection, focused on exploiting the the eyes. Liars always look up and to emotional state and his spoken psychological effort required to control message. There could be many reasons the left. A liar will suffer from dry lips one's emotions while lying.14 The and mouth, and will therefore lick his other than dishonesty for such an underlying theory suggests that it is incongruity, however, such as fear, lips more. Liars blink a lot. Someone more taxing for a person to represent a who is lying will have a facial tick. All nervousness, embarrassment or statement as true if it is, in fact, a lie, uncertainty. of these are regularly listed by people as than if it were actually true. The speaker things to look for when trying to figure has to keep the lie straight and out who is lying.12 A crafty litigator then needs to find consistent, control tone and maintain strategies for converting micro- A person who is motivated to lie, composure. All of the studies cited to expressions into more easily detectable however, will do his very best to this point suggest that liars actually facial cues. That is, the litigator wants control the kinds of behaviors that are achieve these ends quite easily. Ekman to increase the emotional transparency likely to be interpreted by his audience and his team, however, suggested that of the witness. This brings us full circle as signs of dishonesty. In fact, a person there must be ways to detect the stress to the topic covered above. According who is trying to deceive might control on the speaker created by having to to the research of Ekman and others, such mannerisms much more closely maintain a lie. the reason that strategic questioning than someone who knows he has increases transparency, as reported in These researchers discovered that the Levine, et al. article, is that such nothing to hide. As such, focusing on people who are trying to remain stoic the face of the speaker might just be direct and accusatory questioning and unreadable (non-transparent) tend, maximizes the strain on the witness counterproductive for detecting nonetheless, to reveal their true deception. attempting to keep his true emotions in emotional states through their facial check. These tactics increase leakage to expressions, but only for microseconds the point where at least some observers Zuckerman, et al. addressed this 15 concern in a meta-analysis that at a time. That is, we tend to leak our detect the speaker's true emotional reviewed many deception detection true emotions. The problem from a state. studies, focusing on listener deception detection standpoint, performance as a function of what the however, is that these leakages, refer Another method of increasing listener could see and hear.13 The red to as micro-expressions, are so emotional leakage is to force the results were quite intriguing. Being able brief that most people miss them speaker to multitask. The theory completely. underlying this tactic is that increasing

18 ■ Women in the Law ■ February 2017 the cognitive load on the speaker will look at other than the witness's face. so much supporting documentation make it more difficult for him to focus Put a chart on the screen and leave it and testimony from other witnesses. his attention on maintaining his there while questioning the witness on Witnesses are likely to feel substantially composure.18 Vrij and his collaborators the crucial detail about which veracity more vulnerable to lie detection than is discovered that forcing a speaker to is in doubt. Hand out copies of a warranted, but that is a good thing in engage simultaneously in another task document to the jurors while asking the this context. Attorneys also have an while speaking increased the ability of witness the key question. While the opportunity during voir dire to ferret listeners to distinguish true statements jurors are busy passing paper and out those potential jurors who seem from falsehoods. It is also possible to reading the document title, they will overly confident in their own evaluative increase the cognitive load by still be listening to the witness. These capabilities. interrupting the speaker's message and distraction techniques, aimed at the forcing him to stop and restart. listener rather than the speaker, have The research discussed in these been shown to improve deception two articles serves two primary This result is encouraging for detection performance.21 purposes. The first is to tell a litigators who are concerned about cautionary tale about relying on witnesses lying on the stand. In court, III. LESSONS FOR deception detection among jurors to it is wise to make the witness examine EMPLOYMENT LITIGATION win your case. As a general rule, don't a document or other exhibit during do it. The second purpose is to share testimony. Even better, have the The lawyer engaged in an some of the strategies that have been witness come down from the witness employment case faces a troubling uncovered through subsequent stand to demonstrate something to the starting point. These cases very often research that can increase the jury. All of these tactics serve to disrupt involve conflicting accounts of events likelihood that jurors will correctly spot the flow of the witness's narrative, and conversations without much the liars on the witness stand—and which will make it more difficult to supporting documentation for either equally importantly, spot the honest control emotional leakage. position. Only very late in the game witnesses. does an employee (or former C. Redirect isn't only for employee) reflect back on his Through strategic forms of witnesses experience with a company and decide questioning, well-timed presentation of that he was treated in a way that he exhibits and a focus on message While distracting the witness can contends is unlawful. While events content, it is possible to increase the increase his transparency, distraction were transpiring, it never occurs to him likelihood that a jury's verdict will result can also help improve the to keep records or share his stories with from deliberations over the true facts perceptiveness of jurors. When left to others. By the same token, his of the case. their own devices, jurors naturally supervisor hears no complaints and focus on the face of the witness on the feels no need to document any stand, a strategy that we have learned interactions with the employee. It all works to the detriment of deception feels like business as usual. So, when NOTES: detection. How then is an attorney to things turn sour and litigation is in the 1 DePaulo, B. M., Lindsay, J. J., Malone, B. get jurors to focus on the elements of picture, jurors find themselves E., Muhlenbruck, L., Chalrton, K., & Cooper, testimony that more reliably reveal the confronted with a classic he said/she H. “Cues to deception,” Psychological Bulletin, truth of the matter? One of the results said scenario. One party (or both) 129(1), 74-118 (2003). discussed in the first article related to might choose to lie during testimony 2 The companion piece, “He Said What?”: the medium used to convey a message. about certain facts. The litigator would Deception Detection and Employment Research revealed that people more Litigation — Part I: The Gullibe Juror,” like to be able to rely on the common appeared in the February/ March 2014 issue of accurately detected truth in phone calls sense of jurors to separate the truth HR Advisor. 19 than in face-to-face interactions. This from the lies. Alas, research has shown 3 Bond, C. F., & DePaulo, B. M. “Accuracy is consistent with what we have learned time and again that we humans are of deception judgments.” 10 Personality and about the deceptiveness of facial uniformly terrible at evaluating Social Psychology Review, 214–234. (2006) expressions. The key then for the trial whether we are being lied to. 4 See Vrij, A., Granhag, P. A., & Porter, S. attorney is to get jurors to listen to “Pitfalls and Opportunities in Nonverbal and testimony, rather than watch it.20 Employment cases do not take Verbal Lie Detection.” 11 Psychological Science in the Public Interest, 89–121. (2010). place in a vacuum, however. They 5 See Vrij, A., Granhag, P. A., & Porter, S. This provides an opportunity for often raise emotional issues, increasing the strategic use of demonstrative (2010). the likelihood that witness testimony 6 exhibits at trial. If a litigator is Warren, G, Schertler, E and Bull, P., will reflect honest witness attitudes. “Detecting Deception from Emotional and concerned that a witness will attempt to Fear of being “found out” is sufficient Unemotional Cues,” 33 J Nonverbal Behav 59– mislead the jury, that litigator should to deter most witnesses from lying on 69 (2009). provide the jurors with something to the stand, especially in the presence of

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 19 7 George Berkeley, “A Treatise Concerning turns out I was talented (at least at this part of the Principles of Human Knowledge,” 1734. it) but had no confidence in my own ability. section 23. Berkeley's formulation was slightly 18 Vrij, A, Fisher, R., Mann, S. and Leal, S, different. The first appearance of the common “Detecting deception by manipulating cognitive phrasing of the riddle appeared in a physics load.” Trends in Cognitive Sciences, 10 (4). pp. textbook in the early twentieth century: Mann, 141–142 (2006). Charles Riborg and George Ransom Twiss. 19 Physics. Scott, Foresman and Co., 1910, p. 235. See generally, Hancock, J., Thom- Santelli, J., & Ritchie, T. “Deception and design: 8 See Levine, T.R., “A Few Transparent The impact of communication technology on Liars,” Communication Yearbook, 34 (2010) for lying behavior,” In E. Dykstra-Erickson, & M. a review and articulation of the theory that any Tscheligi (Eds.), Proceedings of the 2004 performance above simple chance in truth conference on human factors in computing detection is the function of having a small systems, 129–134. (New York: Association for number of “easy-to-read” liars in the Computing Machinery, 2004), as well as, population. DePaulo, B. M., Kashy, D. A., Kirkendol, S. E., 9 See Levine, T.R., Shaw, A. and Shulman, Wyer, M. M., & Epstein, J. A. “Lying in everyday HC, “Increasing Deception Detection Accuracy life,” 70 Journal of Personality and Social with Strategic Questioning,” 36 Human Psychology, 979–995 (1996). Communication Research, 216-231 (2010) for 20 See Vrij, A., Granhag, P. A., & Porter, S. an excellent treatments of the distinction “Pitfalls and Opportunities in Nonverbal and between demeanor and transparency. Verbal Lie Detection.” 11 Psychological Science 10 Gilovich, T, Savitsky, K, and Medvec, in the Public Interest, 89–121. (2010) V.H., “The Illusion of Transparency: Biased 21 It is important, of course, not to distract Assessments of Others' Ability to Read One's the listener so much that he does not hear the Emotional States,” 75 Journal of Personality words being spoken by the witness. and Social Psychology, 332–346 (1998). Edward P. Schwartz is a 11 Levine, T.R., Shaw, A. and Shulman, HC, “Increasing Deception Detection Accuracy with Consultant in DecisionQuest’s Boston office. Dr. Schwartz provides Strategic Questioning,” 36 Human Communication Research, 216–231 (2010). quantitative and qualitative analysis of pretrial jury behavior – from interviews 12 Vrij, A. (2008). Detecting deceit and lies: Pitfalls and opportunities. West Sussex: Wiley. and focus group studies to mock trials See also Vrij, A., Granhag, P. A., & Porter, S. and large-scale statistical analyses – to (2010) for an excellent treatment of verbal and provide clients with feedback on case non-verbal cues to deception. themes, strategies, evidence, witnesses, 13 Zuckerman, M., DePaulo, B. M., & and presentation style. Further, he Rosenthal, R. (1981). Verbal and nonverbal consults on case evaluation, provides communication of deception. In L. Berkowitz advice on trial strategies, and assists (Ed.), Advances in experimental social with jury selection as well as post- psychology. New York: Academic Press. verdict juror surveys and interviews. Dr. 14 Ekman, P., O'Sulivan, M., & Frank, G. Schwartz is a nationally recognized jury M. (1999). A few can catch a liar. Psychological consultant with excellent analytical Science, 10, 263– 266. acumen and strong market research 15 There are seven universal facial skills who is noted for his ability to expressions that are easily identified by virtually blend the strategic focus of game theory all humans, even babies. They are: fear, joy, and decision theory with the real-world anger, surprise, sadness, contempt and disgust. insights of social psychology to gain a If we could watch speakers in super-slow complete picture of how people absorb, motion, we would all be able to detect their true emotional states without much problem. analyze, and process information. 16 David Matsumoto, Ph.D., Hyi Sung

Hwang, Ph.D., Lisa Skinner, J.D., and Mark Frank, Ph.D., “Evaluating Truthfulness and Detecting Deception,” in FBI Law Enforcement Bulletin (June, 2011).

17 On a personal note, I attended a two hour training program conducted by David

Matsumoto and discovered that I was remarkably proficient at identifying micro- expressions by the end of the session. I had no idea, however, how I was so successful. It felt very much like I was guessing, despite my 83%

accuracy rate. This factoid is interesting in light of the result, reported in the last article, that confidence in deception detection is not at all correlated with actual acumen at doing so. It Reproduced with permission: HR Adviser: Legal and Practical Guidance (Thomson Reuters), Vol. 2 (2), pp. 7-18, March/April 2014

20 ■ Women in the Law ■ February 2017 III. Less Is More? Detecting Lies in Veiled Witnesses

Less Is More? Detecting Lies in Veiled Witnesses

Amy-May Leach, Nawal Ammar, Bennett Kleinberg and Bruno J. Verschuere D. Nicole England, and Laura M. Remigio University of Amsterdam University of Ontario Institute of Technology

Judges in the United States, the United Kingdom, and Canada have ruled that witnesses may not wear the niqab—a type of face veil—when testifying, in part because they believed that it was necessary to see a person’s face to detect deception (Muhammad v. Enterprise Rent-A-Car, 2006; R. v. N. S., 2010; The Queen v. D(R), 2013). In two studies, we used conventional research methods and safeguards to empirically examine the assumption that niqabs interfere with lie detection. Female witnesses were randomly assigned to lie or tell the truth while remaining unveiled or while wearing a hijab (i.e., a head veil) or a niqab (i.e., a face veil). In Study 1, laypersons in Canada (N ϭ 232) were more accurate at detecting deception in witnesses who wore niqabs or hijabs than in those who did not wear veils. Concealing portions of witnesses’ faces led laypersons to change their decision-making strategies without eliciting negative biases. Lie detection results were partially replicated in Study 2, with laypersons in Canada, the United Kingdom, and the Netherlands (N ϭ 291): observers’ performance was better when witnesses wore either niqabs or hijabs than when witnesses did not wear veils. These findings suggest that, contrary to judicial opinion, niqabs do not interfere with—and may, in fact, improve—the ability to detect deception.

Keywords: lie detection, Muslims, witnesses, veiling, minimal information

Wearing a niqab—a veil that covers the wearer’s face, except remove her veil in order to testify (Muhammad v. Enterprise for her eyes—is increasingly prevalent, but contentious. In the Rent-A-Car, 2006). Similarly, in Canada, an alleged victim of 1970s, 1% of the Muslim population wore face veils; currently, childhood sexual assault was ordered to testify at a preliminary approximately one third of female Muslims engage in the practice inquiry without her niqab (R. v. N. S., 2010). Most recently, a (see Figure 1 for sample niqab; al-Ghazali, 2008). There has been defendant who had been charged with witness intimidation was considerable debate about the appropriateness of wearing a niqab directed to remove her niqab while presenting evidence in the (e.g., Khiabany & Williamson, 2008; Mistry, Bhugra, Chaleby, United Kingdom (The Queen v. D(R), 2013). In banning the Khan, & Sauer, 2009; Vakulenko, 2007). In fact, the wearing of a wearing of a niqab while testifying, the various courts attempted to niqab has been officially banned from all public places in several balance the need to establish a witness’s identity, the strength of countries, such as Belgium, Egypt, France, and Turkey (Loi n° the women’s religious beliefs, and the right to freedom of religion. 2010–1192 du 11 octobre 2010 interdisant la dissimulation du Ultimately, however, the right to a fair trial—and the threat to that visage dans l’espace public, 2010; Loi visant a` interdire le port de right posed by allowing a witness to wear the niqab while testify- tout vêtement cachant totalement ou de manière principale le ing—appeared to override the witnesses’ right to veil. The presid- visage, 2011; Syed, 2010). ing judges opined that, in an adversarial trial, a judge must be able The permissibility of the niqab has also been called into ques- to see a witness’s face to assess her truthfulness. tion by the courts. For example, a judge in an American small Considerable psychology–law research has been devoted to claims court dismissed a plaintiff’s complaint when she refused to testing assumptions underlying legal decisions and laws. For example, Wells and Quinlivan (2009) found that beliefs about human cognition, which formed the basis of the U.S. Supreme Court’s This article was published Online First June 27, 2016. decision on how to evaluate claims of suggestiveness of police Amy-May Leach, Nawal Ammar, D. Nicole England, and Laura M. lineups in Manson v. Braithwaite (1977), were inconsistent with Remigio, Faculty of Social Science and Humanities, University of Ontario contemporary research findings in the eyewitness identification Institute of Technology; Bennett Kleinberg and Bruno J. Verschuere, literature. In the current studies, we examined the notion embodied Clinical Psychology, University of Amsterdam. in important court decisions in the United States, the United This research was supported by Social Sciences and Humanities Re- Kingdom, and Canada (e.g., N. S. v. Her Majesty the Queen et al., search Council of Canada Grant 430-2011-0407 to Amy-May Leach and 2012): that a fact-finder’s ability to detect deception among wit- Nawal Ammar. We also thank Brian Cutler for his comments on the nesses is compromised by the niqab. manuscript. Correspondence concerning this article should be addressed to Amy- Typically, observers’ lie detection performance is poor. Average May Leach, University of Ontario Institute of Technology, Faculty of accuracy for laypersons and justice officials is very close to 50%, Social Science and Humanities, 2000 Simcoe Street North, Oshawa, On- or chance levels (Aamodt & Custer, 2006; Bond & DePaulo, tario, Canada. E-mail: [email protected] 2006). In the majority of lie detection studies, however, lie-tellers’

401

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 21 402 LEACH ET AL.

Ambady & Rosenthal, 1992). Focusing on general impressions might discourage the use of irrelevant details and increase effi- cient, intuitive processing without taxing cognitive resources (Am- bady, 2010; Murphy & Balzer, 1986). Minimization of information principles have been applied to lie detection. In one direct test, observers who were afforded brief glimpses of behavior were more accurate than observers who viewed lie- and truth-tellers’ entire accounts (Albrechtsen, Meiss- ner, & Susa, 2009). It is not necessarily that this approach encour- ages unconscious decision-making, but rather that it focuses observers on a limited number of diagnostic cues (Street & Richardson, 2014). These minimization effects have also been examined in terms of the medium of presentation (i.e., audio vs. visual vs. audiovisual; Burgoon, Blair, & Strom, 2008; Zucker- man, Koestner, & Colella, 1985). This research might be the most relevant to the study of face veiling because it involves restricting the nonverbal and verbal cues that are available for decision- making. A meta-analysis of 50 studies (Bond & DePaulo, 2006) revealed that overall lie detection accuracy was similar, whether observers received audio (i.e., more restricted) or audiovisual (i.e., less restricted) information. To date, there has been no research on how exposure to the full range of verbal cues, but only a subset of nonverbal cues, affects performance. We found no empirical evidence in the lie detection literature suggesting that a niqab should impair lie detection because it Figure 1. One type of veil—the niqab—that is the focus of this research conceals portions of the wearers’ face; rather, existing research project. suggests that the opposite could occur. Niqabs should minimize the amount of information that is available to observers and prevent them from basing their lie detection decisions on misleading facial and truth-tellers’ faces were visible. Our literature review uncov- cues (e.g., smiling; DePaulo et al., 2003). In turn, the veiling of the ered no previously published research on the effects of religious witness might force observers to attend to sources of information garments on lie detection. that are more diagnostic of deception, such as verbal content (Vrij, A few studies were indirectly relevant to peoples’ abilities to 2008). Niqabs also explicitly highlight a specific subset of non- detect deception among veiled witnesses. Research on cross- verbal cues (i.e., the witnesses’ eyes). It is widely believed that a cultural lie detection was informative, for example. It is highly person’s eyes reveal deception (The Global Deception Research unlikely that every single defendant, plaintiff, witness, and Team, 2006), and the eye region can be used to identify complex decision-maker (e.g., juror, judge) involved in a case would wear mental states (e.g., guilt; Baron-Cohen et al., 1997). In particular, a niqab. Although viewing veiled witnesses would not constitute blinking, and pupil dilation are effective cues to deceit in certain cross-cultural lie detection, observers’ inexperience with veiling contexts (e.g., DePaulo et al., 2003; Leal & Vrij, 2008). By might be analogous to a lack of familiarity between cultures. In encouraging the use of verbal cues and/or eye region cues, niqabs two studies, laypersons were slightly better at detecting deceit could actually facilitate the detection of deception. within their own cultures than across cultures (Bond & Atoum, Although niqabs may lead to improved lie detection perfor- 2000; Bond, Omar, Mahmoud, & Bonser, 1990). Yet, Vrij and mance, they might also elicit response biases. Veiled Muslim Winkel (1991) failed to find significant differences between Dutch women report being stared at, insulted, and assaulted (Clarke, and Surinamese observers’ cross-cultural lie detection. Given the 2013). Being Muslim is associated with an aggressive stereotype limited number of studies and mixed findings (see Taylor, Larner, (Fischer, Greitemeyer, & Kastenmuller, 2007), and niqabs are Conchie, & van der Zee, 2014, for a full discussion), it remains particularly threatening (Behiery, 2013). It is possible that people unclear from this literature whether there is a meaningful disad- attribute a range of other negative behaviors, such as lying, to vantage to detecting deception in witnesses in niqabs. women who wear veils (see Hoodfar, 1997 for a similar argument Instead, it might be important to focus on the more limited about head veiling). The typical dark color of niqabs could also information that is afforded by niqabs as compared to bare faces. invoke the black clothing stereotype, in which individuals who People are able to make inferences based on minimal information. wear dark (vs. light) colors are less likely to be judged as credible For example, point-light displays of biological motion and static (Akehurst, Kohnken, Vrij, & Bull, 1996). In turn, there might be a images of facial features can be used to discern others’ attributes tendency to label women who wear niqabs as lie-tellers, regardless (Baron-Cohen, Wheelwright, & Jolliffe, 1997; Blais, Roy, Fiset, of the underlying veracity of those witnesses’ accounts. Maeder, Arguin, & Gosselin, 2012; Troje, 2002). Similarly, another form of Dempsey, and Pozzulo (2012) examined whether an alleged sexual minimization, thin slices (i.e., exposure to less than 5 minutes of assault victim’s veiling influenced the perceived culpability of the behavior) reveal performance, interpersonal relationships, and in- defendant. In their study, the victim was described as wearing a dividual differences (e.g., Ambady, Bernieri, & Richeson, 2000; burqa (i.e., a garment in a solid color that, in addition to covering

22 ■ Women in the Law ■ February 2017 VEILED WITNESSES 403 the hair and face, conceals the entire body), a hijab, or no veil. minutes to prepare their testimony and, as in real trials, they were Mock jurors, who read transcripts of the victim’s testimony, were provided with the questions that would be asked by the defense more confident in the defendant’s guilt when the victim was lawyer. Once they were prepared, witnesses were randomly as- described as wearing a burqa or hijab than when she did not veil. signed to don a black niqab, a black hijab, or remain unveiled. In In sum, various sources directly and indirectly associated with addition, they were asked to wear an opaque black shawl to veiling led us to examine the possibility that niqabs produced a conceal and control for clothing. Veils and shawls were placed on response bias during lie detection. No research, to date, has ex- the witnesses by a trained research assistant. amined the effects of actively minimizing only a subset of non- Witnesses were interviewed by two female experimenters. To verbal cues—while highlighting others—on response bias. simulate courtroom procedures, one experimenter played the role of the sympathetic defense lawyer and asked 16 information- gathering questions (e.g., “Please describe everything that you saw Study 1 the woman do.”). The other experimenter conducted a challenging In Study 1, we examined participants’ lie detection accuracy, cross-examination as the prosecutor and asked seven unanticipated response biases, and decision strategies when evaluating the tes- questions (e.g., “The police found the man’s laptop. The defen- timony of eyewitnesses in three veiling conditions: niqab, hijab, dant’s fingerprints were on it. How do you explain that?”). Role and no veil. We hypothesized that lie detection accuracy would be assignment was counterbalanced, and both experimenters were higher in the niqab condition than in the hijab or no-veil conditions blind to the veracity of the witness’s testimony. To increase the because it would minimize the availability of misleading cues to stakes associated with deception, witnesses were told that they deception and facilitate the use of more effective strategies (e.g., might receive $50 if they convinced both experimenters that they Albrechtsen et al., 2009). In addition, we predicted that veils were telling the truth. In fact, all of the witnesses were given the would activate stereotypically negative views of Muslim women opportunity to win the incentive in a draw. At the end of each (e.g., Fischer et al., 2007); therefore, we expected a lie bias (i.e., session, witnesses rated their perceptions of their interviews. These tendency to indicate that witnesses were lying) in the niqab and data are available from the corresponding author. hijab conditions but not in the no-veil condition. Given that niqabs The interview was the only portion of the session that was are portrayed less positively than hijabs (Behiery, 2013), we hy- videotaped. We excluded data from 19 witnesses because the pothesized that the lie bias would be stronger in the niqab condi- quality of the video footage was poor or their garments (i.e., veils tion than in the hijab condition. Finally, we conducted exploratory or shawls) were askew. In addition, one witness in the hijab analyses to determine whether expected lie detection effects could condition confessed to having seen the woman steal items (i.e., she be accounted for by participants in the niqab condition attending to did not follow our instruction to lie about the theft). Clips were not witnesses’ eyes and the content of their accounts (i.e., verbal cues) selected based on the cues revealed by witnesses. The full range of to a greater extent than participants who were able to see wit- lie-telling and truth-telling proficiency is present in the justice nesses’ entire faces (i.e., the hijab and no-veil conditions) and/or system (i.e., there is a continuum from poor to proficient truth- the witnesses’ actual nonverbal and verbal behaviors. telling, and poor to proficient lie-telling). Generalizability was ensured by randomly assigning witnesses to condition and presenting all of their responses to observers, regardless of their quality. Method In total, we compiled clips of 10 lie-tellers and 10 truth-tellers in Participants. Two hundred and 32 students at a Canadian each veiling condition (M length per interview ϭ 2.06 min, SD ϭ university (138 females, 94 males; M age ϭ 20.09 years, SD ϭ 0.37). Demographic characteristics of the witnesses were similar 3.83) completed the study in exchange for course credit. Partici- across conditions. Clip order was randomly assigned and counter- pants self-identified as belonging to the following ethnic groups: balanced within each condition, producing two versions of each set Arab/West Asian (n ϭ 22), Black (n ϭ 25), Chinese (n ϭ 8), of 20 videos. White (n ϭ 74), Hispanic (n ϭ 1), Korean (n ϭ 1), Latin American Coding. All of the videos were coded for the onsets and (n ϭ 3), South Asian (n ϭ 79), South East Asian (n ϭ 10), other offsets of nonverbal and verbal cues using Datavyu (i.e., a video (n ϭ 9). The majority of participants (n ϭ 223) did not wear hijabs coding and data visualization tool; Datavyu Team, 2014). In ad- or niqabs and self-identified as Christians (n ϭ 95). dition, research assistants coded the degree to which certain cues Study design. We employed a 2 (veracity: lie-tellers vs. truth- occurred (see AppendixϾ). One research assistant coded all of the tellers) ϫ 3 (veiling condition: niqab vs. hijab vs. no veil) mixed- footage, whereas another research assistant coded 25% of each factors design. We manipulated veiling condition between partic- video, as recommended by Datavyu. Interrater reliability, calcu- ipants to decrease the potential impact of demand characteristics. lated using ICCs, was high (M ϭ .86, SD ϭ .16). Materials. Lie detection measure. Participants were asked to indicate Video footage. In individual sessions, female witnesses (N ϭ whether the witness in the video clip was lying or telling the truth 80, M age ϭ 20.23 years, SD ϭ 5.74) were shown a video of a about having seen the woman steal the items. Participants were woman who was watching a stranger’s bag. As determined by awarded a “1” for each correct response and a “0” for each random assignment, half of the women also observed her stealing incorrect response. Then, all scores were averaged to determine items from the bag. Then, all of the witnesses were informed that overall accuracy, resulting in a score between 0 (no lie detection the woman had been accused of theft and they were being called ability)and1(perfect lie detection ability). to testify on her behalf (i.e., they were to state that they did not see Confidence measure. Using a scale, from 0% (not at all her steal anything). Thus, half of the witnesses were lying and half confident) to 100% (extremely confident), participants indicated of the witnesses were telling the truth. Witnesses were given 2 how confident they were in each lie detection judgment. Ratings

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 23 404 LEACH ET AL. were averaged across witnesses to yield overall confidence scores. Post hoc tests, using Tukey’s honest significant difference, re- Because confidence analyses were exploratory in nature, and they vealed that participants were more accurate when viewing wit- did not reveal significant effects, we will not report them here. nesses who wore hijabs or niqabs than those who did not wear Interested readers can obtain the data from the corresponding veils, p Ͻ .001, d ϭ 0.63, 95% confidence interval (CI) [0.30, author. 0.96], and p ϭ .038, d ϭ 0.33, 95% CI [0.01, 0.65], respectively. Cue use measure. Participants were asked to indicate which There was no significant difference, in terms of overall accuracy, verbal cues (e.g., amount of detail) and nonverbal cues (e.g., eye between participants in the hijab and niqab conditions, p ϭ .175, contact) they used to make their decisions from the same list d ϭ 0.32, 95% CI [Ϫ0.00, 0.63]. Regardless of veiling condition, containing empirically verified actual and perceived cues to de- participants were more accurate when judging truth-tellers (M ϭ ception that was coded by research assistants. For each cue, .72, SD ϭ .20) than lie-tellers (M ϭ .38, SD ϭ .21), F(1, 229) ϭ 2 participants were given a “1” if they indicated that they had used 210.24, p Ͻ .001, ␩p ϭ .48, d ϭ 1.66, 95% CI [1.45, 1.87]. that cue to make their decisions, and a “0” if they had not used the However, there was no significant interaction between veracity 2 cue. Each variable was classified as a nonverbal or verbal cue. and veiling condition, F(2, 229) ϭ 1.19, p ϭ .306, ␩p ϭ .01. Then, we calculated a difference score (i.e., subtracted overall Participants’ signal detection. As noted by Meissner and verbal from nonverbal cue use). Kassin (2002), focusing solely on accuracy can obscure the dis- Experience measure. We asked participants to report their tinction between discrimination (i.e., the ability to identify lie- and experiences with lie detection on a scale from 1 (not at all)to5 truth-tellers) and response bias (i.e., the tendency to choose a (extremely) and describe any relevant additional experience in the particular response). Given our interest in both of these factors, we area. Participants also indicated whether they had ever worked in followed Meissner and Kassin’s example and conducted a signal law enforcement and had taken any courses related to lie detection. detection analysis. All calculations were based on Wixted and Procedure. Given that the two variables were manipulated in Lee’s (2014) formulas. Specifically, we calculated “hits” (i.e., the the videotaped stimuli, the procedure was identical for all partic- percentage of correct classifications of lie-tellers) and “false ipants. Individually, following random assignment to one of the alarms” (i.e., the percentage of truth-tellers incorrectly classified as three veiling conditions, we showed each participant 20 video clips lie-tellers). of lying and truth-telling witnesses using a computer program Discrimination. We conducted a one-way ANOVA, with (Jarvis, 2008). After each clip, the participant rendered a lie veiling condition as the independent variable, on discrimination detection and confidence rating. At the end of the session, the (i.e., d ). Echoing the overall accuracy analysis, there was a sig- participant indicated the cues that he or she had used to make = nificant effect of veiling, F(2, 229) ϭ 8.18, p Ͻ .001, ␩2 ϭ .07 (see decisions and provided demographic information. Each session p Table 1). Post hoc tests revealed that participants were better able lasted approximately one hour. to discriminate between lie-tellers and truth-tellers in hijabs than those who did not wear veils, p Ͻ .001, d ϭ 0.63, 95% CI [0.30, Results 0.95]. The difference between participants in the niqab and no-veil Preliminary analyses revealed nonsignificant effects of partici- conditions approached significance, p ϭ .056, d ϭ 0.38, 95% CI pant gender, race, veiling, religious affiliation, and lie detection [0.06, 0.70]. Performance was similar when participants viewed experience. All reported analyses are collapsed across those fac- witnesses wearing hijabs or niqabs, p ϭ .196, d ϭ 0.26, 95% CI tors. [Ϫ0.05, 0.58]. Participants’ accuracy. We conducted a Veracity ϫ Veiling We also compared participants’ d scores to zero (i.e., no sen- = Condition analysis of variance (ANOVA) on accuracy scores. The sitivity). Participants could discriminate between lie- and truth- cell means are shown in Table 1. There was a significant main telling witnesses who wore niqabs, t(77) ϭ 5.18, p Ͻ .001, d ϭ 2 effect of veiling condition, F(2, 229) ϭ 9.07, p Ͻ .001, ␩p ϭ .07. 0.59, 95% CI [0.34, 0.83], or hijabs, t(76) ϭ 6.84, p Ͻ .001, d ϭ

Table 1 Lie Detection Performance

Discrimination (d ) Response bias (␤) Veiling Accuracy = condition M (SD) M (SD) 0 (no sensitivity) M (SD)1a (no bias)

Study 1 Niqab .55a (.09) .23ab (.39) Ͼ .94a (.31) — Hijab .58a (.10) .34a (.44) Ͼ .86a (.37) Ͻ No veil .52b (.09) .08b (.39) — .99a (.38) — Study 2 Niqab .57a (.12) .30a (.52) Ͼ .93ab (.31) Ͻ Hijab .59a (.11) .43a (.47) Ͼ .86a (.26) Ͻ No veil .51b (.13) .03b (.39) — 1.01b (.41) — Note. Means sharing a common subscript are not statistically different at ␣ϭ.05 according to Tukey’s honest significant difference tests. Dashes in column 4 indicate that there was no difference from 0 and dashes in column 6 indicate that there was no difference from 1. a Means significantly less than 1 indicate a truth bias.

24 ■ Women in the Law ■ February 2017 VEILED WITNESSES 405

0.78, 95% CI [0.39, 1.17]; however, they were unable to do so in difference score (i.e., subtracted overall verbal from nonverbal cue the no-veil condition, t(76) ϭ 1.70, p ϭ .094, d ϭ 0.19, 95% CI use). A Veracity ϫ Veiling Condition ANOVA on the difference [Ϫ0.97, 0.29]. score data revealed a significant main effect of veiling, F(1, 54) ϭ 2 Response bias. A one-way ANOVA revealed that partici- 8.15, p ϭ .001, ␩p ϭ .23. Witnesses who wore niqabs (M ϭ .32, pants’ response biases (i.e., ␤) were not affected by veiling con- SD ϭ .36) were more likely to reveal verbal (vs. nonverbal) 2 dition, F(2, 229) ϭ 2.10, p ϭ .126, ␩p ϭ .02 (see Table 1). By information than witnesses who wore hijabs (M ϭϪ.26, SD ϭ comparing ␤ scores to one (i.e., no response bias), we could .60), p ϭ .001, d ϭ 1.18, 95% CI [0.49, 1.87], or did not veil examine participants’ tendencies to label witnesses as lie-tellers or (M ϭϪ.06, SD ϭ .36), p ϭ .031, d ϭ 1.06, 95% CI [0.38, 1.75]. truth-tellers. Participants exhibited a truth bias toward witnesses There were no differences between witnesses who wore hijabs or who were wearing hijabs, t(76) ϭϪ3.23, p ϭ .002, d ϭϪ0.37, did not veil, p ϭ 1.000, d ϭ 0.00, 95% CI [Ϫ0.31, 0.32]. There 95% CI [Ϫ0.60, Ϫ0.14]. There was no evidence of bias in the were no other significant effects. niqab condition, t(77) ϭϪ1.66, p ϭ .101, d ϭϪ0.19, 95% CI We also conducted an exploratory Veracity ϫ Veiling Condi- [Ϫ0.28, Ϫ0.09] or in the no-veil condition, t(76) ϭϪ0.30, p ϭ tion MANOVA on the overall presence of empirically verified .767, d ϭϪ0.03, 95% CI [Ϫ0.05, Ϫ0.02]. cues to deception. Participants rated both diagnostic and nondiag- Participants’ cue use. Although not a primary research ques- nostic cues to reduce any effects of demand characteristics; only tion, we were interested in whether participants’ lie detection the former were analyzed here. Including known nondiagnostic performance could be explained by the cues that they reported cues in the analysis would have unnecessarily impeded the likeli- using to render decisions. We conducted a multivariate analysis of hood of uncovering significant effects. Interested readers can ob- variance (MANOVA) on participants’ self-reported reliance on the tain these analyses from the corresponding author, however. In eye region (i.e., blinking, eye contact, and pupil dilation) to detect total, we analyzed 15 cues (i.e., fidgeting, inconsistencies, admit- deception. There was only a significant effect of veiling condition ted lack of memory, length of response, negative statements, on the combined dependent variables, F(6, 456) ϭ 2.56, p ϭ .019; spontaneous corrections, unfriendly facial expressions, word/ 2 Pillai’s trace ϭ .07 ␩p ϭ .03. Examining the univariate effects phrase repetitions, vocal tension, coherence/plausibility, vagueness, revealed that veiling did not affect the use of eye contact, F(2, cooperativeness, nervousness, amount of detail, and pitch). There was 2 229) ϭ 0.17, p ϭ .848, ␩p ϭ .00, or pupil dilation, F(2, 229) ϭ no significant interaction between veracity and veiling condition, 2 2 0.99, p ϭ .373, ␩p ϭ .01, as cues to deceit. In fact, eye contact was F(28, 84) ϭ 0.88, p ϭ .639; Pillai’s trace ϭ .45 ␩p ϭ .23, nor a frequently cited as a cue to deception across all conditions (M ϭ significant main effect of veiling on the combined dependent vari- 2 .92, SD ϭ .27). However, blinking use did vary with veiling ables, F(28, 84) ϭ 1.02, p ϭ .449; Pillai’s trace ϭ .51 ␩p ϭ .25. 2 condition, F(2, 229) ϭ 1.44, p ϭ .003, ␩p ϭ .05. Post hoc tests However, there was a statistically significant difference between indicated that participants were equally likely to report that they lie-tellers and truth-tellers, F(14, 41) ϭ 2.46, p ϭ .013; Pillai’s 2 used blinking to detect deception when the witnesses wore niqabs trace ϭ .46 ␩p ϭ .46. A closer examination of the univariate effects (M ϭ .60, SD ϭ .49), or did not veil (M ϭ .60, SD ϭ .49), p ϭ revealed that lie-tellers spoke in a higher pitch, were less cooperative, .998, d ϭ 0.00, 95% CI [Ϫ0.32, 0.32]. Participants stated that they and provided accounts that were less coherent than truth-tellers (see relied less on witnesses’ blinking in the hijab condition (M ϭ .36, Table 2). SD ϭ .48) than in the niqab, p ϭ .008, d ϭϪ0.49, 95% CI [Ϫ0.82, Ϫ0.17], or no-veil conditions, p ϭ .010, d ϭϪ0.49, 95% Discussion CI [Ϫ0.82, Ϫ0.17]. Veiling also affected overall reported cue use, F(2, 229) ϭ As predicted, participants were more accurate when witnesses 2 14.75, p Ͻ .001, ␩p ϭ .11. Participants were more likely to state wore niqabs than when witnesses did not wear veils. They did not, that they based their decisions on verbal cues than nonverbal cues however, exhibit different response biases toward the former when witnesses wore niqabs (M ϭϪ.17, SD ϭ .10) than when group. Our sample consisted of participants from an ethnically and they wore hijabs (M ϭϪ.10, SD ϭ .08), d ϭϪ0.77, 95% CI religiously diverse student population at a Canadian university. [Ϫ1.10, Ϫ0.44], or did not veil (M ϭϪ.10, SD ϭ .08), Perhaps the characteristics of the student body, exposure to a d ϭϪ0.77, 95% CI [Ϫ1.10, Ϫ0.44], all ps Ͻ .001. There were no cross-cultural curriculum, and/or social desirability concerns could differences, in terms of overall cue use, when participants viewed account for the lack of response bias toward witnesses wearing witnesses who wore niqabs or did not veil, p ϭ 1.000, d ϭ 0.00, veils. Indeed, the demographic composition and diversity of reli- 95% CI [Ϫ0.31, 0.32]. gious life in Canada, as well as the country’s historical endorse- Coded cues. We performed the same analyses on the eye ment of a cultural mosaic approach to multiculturalism, might have region as above to allow for a comparison between participants’ made biased decision-making unlikely. It may be more prevalent self-reports and the actual presence of cues to deception. Only the in geographical regions where the niqab has been more publically effect of veracity was statistically significant, F(2, 53) ϭ 3.84, p ϭ opposed and expressing a negative response bias would be more 2 .028; Pillai’s trace ϭ .13 ␩p ϭ .67. Veracity did not affect blinking, socially acceptable. 2 F(1, 54) ϭ 0.06, p ϭ .812, ␩p ϭ .06. However, lie-tellers (M ϭ The lack of bias in the current study could also be attributed to 26.83, SD ϭ 10.81) made eye contact less frequently than truth- the context in which data collection took place. R. v. N. S. (2010), 2 tellers (M ϭ 34.20, SD ϭ 10.72), F(1, 54) ϭ 7.67, p ϭ .008, ␩p ϭ the Canadian case in which a witness was asked to remove her veil .78, d ϭϪ0.68, 95% CI [Ϫ1.21, Ϫ0.15]. in court, was tried within the university’s catchment area. There To determine whether one type of cue was more likely to occur, were numerous appeals and a Supreme Court trial that took place the data were transformed into z scores and each variable was in the midst of data collection (see N. S. v. Her Majesty the Queen classified as a nonverbal or verbal cue. Then, we calculated a et al., 2012). Simultaneously, a neighboring provincial govern-

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 25 406 LEACH ET AL.

Table 2 Mean Nonverbal and Verbal Behaviors by Veracity

Lie-tellers Truth-tellers Behaviors M (SD) M (SD) d [CI] p

Unfriendly facial expressions 0.90 (1.24) 0.70 (1.75) .13 [Ϫ.38, .65] .600 Fidgeting 1.47 (2.64) 0.60 (1.25) .42 [.07, .52] .111 Overall nervousness 3.87 (1.01) 3.73 (0.83) .15 [.06, .52] .557 Word or phrase repetitions 0.67 (1.12) .53 (1.33) .11 [.06, .52] .680 Pitch 3.13 (0.35) 2.77 (0.54) .79 [.25, 1.33] .001 Vocal tension 2.07 (1.02) 1.93 (0.87) .15 [Ϫ.37, .67] .559 Length of responses 23.60 (0.77) 23.43 (0.94) .20 [Ϫ.32, .71] .467 Coherence 4.73 (0.52) 5.00 (0.00) Ϫ.73 [Ϫ1.27,Ϫ.20] .008 Amount of detail 2.60 (1.10) 2.70 (0.92) Ϫ.09 [Ϫ.61, .41] .706 Spontaneous corrections 1.40 (1.38) 1.23 (1.33) .13 [Ϫ.39, .64] .633 Admitted lack of memory 0.27 (0.83) 0.07 (0.37) .31 [Ϫ.21, .83] .236 Inconsistencies 0.07 (0.25) 0.00 (0.00) .40 [Ϫ.12, .92] .139 Vagueness 2.63 (1.27) 2.90 (1.35) Ϫ.21 [Ϫ.72, .31] .437 Negative statements 0.00 (0.00) 0.00 (0.00) Cooperativeness 4.87 (0.35) 5.00 (0.00) Ϫ.53 [Ϫ1.05,Ϫ.00] .044 Note. CI ϭ confidence interval.

ment drafted Bill 60 (2013), which limited State employees’ abil- Method ities to wear overt religious symbols and conceal their faces; in essence, it would have severely restricted wearing of the niqab. In Participants. Two hundred and 91 students at universities in response, the university’s local hospital mounted a recruitment Canada, the United Kingdom, and the Netherlands (201 females, campaign—including ads and signs on city streets—depicting a 90 males; M age ϭ 21.11 years, SD ϭ 4.33) completed the study medical professional wearing a hijab next to the slogan, “We don’t in exchange for extra credit or a small honorarium. Participants care what’s on your head. We care what’s in it” (Mok, 2013). self-identified as belonging to the following ethnic groups: Arab/ Thus, participants were exposed to explicit messages from local West Asian (n ϭ 19), Black (n ϭ 20), Chinese (n ϭ 7), White (n ϭ authorities that veiling should not bias decision-making, in addi- 194), Hispanic (n ϭ 1), Korean (n ϭ 1), Latin American (n ϭ 1), tion to significant public debate surrounding the permissibility of South Asian (n ϭ 26), South East Asian (n ϭ 13), Other (n ϭ 9). wearing niqabs in court. That exposure might have affected any The majority of participants (n ϭ 286) did not report wearing any preexisting response biases. type of veil or having a religious affiliation (n ϭ 171). Study design. We employed a 2 (veracity: lie-tellers vs. truth- tellers) ϫ 3 (country: Canada vs. the Netherlands vs. the United Study 2 Kingdom) ϫ 3 (veiling condition: niqab vs. hijab vs. no veil) Study 2 served as both a direct replication (in Canada) as well mixed-factors design. As in Study 1, veiling condition was a as an extension to two other countries (i.e., the Netherlands and the between-participants factor, whereas veracity was a within- United Kingdom). We chose the Netherlands because its govern- participants factor. ment recently came very close to banning the niqab (Government Materials and procedure. The procedure was similar to of the Netherlands, 2012). We also sought to replicate findings in Study 1, with a few key differences. Participants in the United the United Kingdom because a ruling on the permissibility of Kingdom and the Netherlands were not asked to provide the cues wearing a niqab in British courts was imminent. Indeed, shortly that they used to render their lie detection decisions. In addition, after data collection began, a judge ruled that a woman must unveil we assessed the English proficiency of participants in the Nether- in court (The Queen v. D(R), 2013). A comparison between these lands, using the criteria established by the Centre for Canadian locations and Canada would further test the generalizability of Language Benchmarks (2010), to ensure that they could under- results. stand the witnesses’ accounts. Dutch participants were asked to As in Study 1, we expected that participants would be better able self-report their overall English proficiency on a 12-point scale to detect deception in witnesses who wore niqabs than in witnesses (Basic ϭ 1–3; Intermediate ϭ 4–8; Advanced ϭ 9–12). Average who did not wear veils. We hypothesized that response bias would proficiency was on the boundary between Intermediate and Ad- vary by region. Specifically, participants in Canada were expected vanced (M ϭ 8.87, SD ϭ 1.88). At the conclusion of each session, to exhibit similar response tendencies, regardless of veiling con- participants listened to two messages that were read aloud in dition, replicating the findings from the first study. We posited that English. After each message, they were asked three multiple- the participants in the Netherlands would exhibit the originally choice questions about its content. Each correct answer was hypothesized pattern of response due to the government’s stance awarded a “1,” whereas each incorrect answer was awarded as “0.” on veiling. Dutch participants were expected to be more likely to Thus, the highest possible score was 6 (out of 6 questions). indicate that women were lying when they were wearing niqabs Participants’ objective language comprehension was extremely than when they did not wear veils. We did not have set hypotheses high (M ϭ 5.02, SD ϭ 0.95) and would be considered “Advanced” about the nature of response bias in the U.K. sample. according to the Canadian Language Benchmarks. Our ANOVA

26 ■ Women in the Law ■ February 2017 VEILED WITNESSES 407 revealed that there was a similar distribution of English compre- 0.24, 95% CI [Ϫ0.04, 0.53], or without veils, p ϭ .198, hension scores across veiling conditions, F(1, 96) ϭ 1.94, p ϭ d ϭϪ0.22, 95% CI [Ϫ0.50, 0.06]. There was no significant main 2 2 .150, ␩p ϭ .04. effect of country, F(2, 281) ϭ 0.97, p ϭ .382, ␩p ϭ .01, or interaction between the variables, F(4, 281) ϭ 0.44, p ϭ .778, ␩2 ϭ .01. Results p We compared participants’ ␤ scores to one (i.e., no bias) to There were nonsignificant effects of race, gender, veiling, reli- examine their tendencies to label witnesses as lie-tellers or truth- gious affiliation, and lie detection experience. Thus, we collapsed tellers within each veiling condition. Participants exhibited a truth across those variables when conducting the following analyses. bias toward witnesses in niqabs, t(95) ϭϪ2.27, p ϭ .025, Participants’ accuracy. A Veracity ϫ Country ϫ Veiling d ϭϪ0.23, 95% CI [-[Ϫ0.43, Ϫ0.02] and hijabs, t(97) ϭϪ5.21, Condition ANOVA indicated that there was a significant main p Ͻ .001, d ϭϪ0.53, 95% CI [Ϫ0.79, Ϫ0.26]. They did not 2 effect of veiling condition, F(2, 281) ϭ 13.28, p Ͻ .001, ␩p ϭ .09 exhibit response biases when witnesses did not wear veils, t(96) ϭ (see Table 1). Post hoc tests revealed that participants were better 0.29, p ϭ .775, d ϭ 0.03, 95% CI [0.01, 0.04]. able to detect the deception of women who wore niqabs or hijabs than of those who did not veil, p ϭ .001, d ϭ 0.48, 95% CI [0.19, Discussion 0.77] and p ϭ .001, d ϭ 0.66, 95% CI [0.37, 0.95], respectively. Performance in the niqab and hijab conditions was similar, p ϭ We partially replicated Study 1’s primary findings. Participants .392, d ϭϪ0.17, 95% CI [Ϫ0.46, 0.11]. In addition, participants were more accurate at detecting the deception of witnesses who were more accurate when judging truth-tellers (M ϭ .71, SD ϭ wore niqabs or hijabs than that of witnesses who did not wear .22) than lie-tellers (M ϭ .39, SD ϭ .20), F(1, 281) ϭ 225.14, p Ͻ veils. There was no evidence of a negative response bias toward 2 .001, ␩p ϭ .45, d ϭ 1.52, 95% CI [0.18, 1.34]. There was no women who veiled in any country. Rather, participants exhibited a significant main effect of country, F(2, 281) ϭ 1.44, p ϭ .240, tendency to indicate that women who wore niqabs or hijabs were 2 ␩p ϭ .01. Interactions between veracity and veiling condition, F(2, telling the truth. 2 281) ϭ 0.13, p ϭ .878, ␩p ϭ .00, veracity and country, F(2, 281) ϭ 2 2.70, p ϭ .069, ␩p ϭ .02, country and veiling condition, F(2, 2 General Discussion 281) ϭ 0.88, p ϭ .475, ␩p ϭ .01, and all three variables, F(4, 2 281) ϭ 1.06, p ϭ .376, ␩p ϭ .02 were also nonsignificant. Contrary to the assumptions underlying the court decisions cited Participants’ signal detection. As in Study 1, we used a earlier, lie detection was not hampered by veiling across two signal detection analysis to examine the independent contributions studies. In fact, observers were more accurate at detecting decep- of discrimination and bias. tion in witnesses who wore niqabs or hijabs than those who did not Discrimination. We performed a Country ϫ Veiling Condi- veil. Discrimination between lie- and truth-tellers was no better tion ANOVA on discrimination (i.e., d ). Again, there was a than guessing in the latter group, replicating previous findings = significant effect of veiling condition, F(2, 281) ϭ 14.37, p Ͻ (Bond & DePaulo, 2006). It was only when witnesses wore veils 2 .001, ␩p ϭ .09 (see Table 1). Post hoc tests indicated that partic- (i.e., hijabs or niqabs) that observers performed above chance ipants were better able to discriminate between lie-tellers and levels. Thus, veiling actually improved lie detection (see Table 1). truth-tellers in niqabs and hijabs than those who did not wear veils, It is unlikely that these findings were simply false positives. p ϭ .001, d ϭ 0.59, 95% CI [0.29, 0.88] and p Ͻ .001, d ϭ 0.96, Simmons, Nelson, and Simonsohn (2011) have identified four 95% CI [0.63, 1.22], respectively. Participants performed similarly researcher degrees of freedom that can increase Type I error: when viewing witnesses who were wearing hijabs or niqabs, p ϭ disclosing only certain subsets of conditions or dependent vari- .232, d ϭϪ0.26, 95% CI [Ϫ0.54, Ϫ0.02]. There was no signifi- ables, employing covariates, and altering the sample size. We did 2 cant main effect of country, F(2, 281) ϭ 0.86, p ϭ .424, ␩p ϭ .01, not engage in any of those practices. All conditions and dependent or interaction between the variables, F(4, 281) ϭ 0.93, p ϭ .444, variables were reported, and covariates were not used. The sample 2 ␩p ϭ .01. sizes differed between the two studies, but the difference was not One-sample t tests, comparing discrimination scores to zero due to an attempt to manipulate significance. Rather, because this (i.e., no sensitivity), revealed that participants could discriminate work was the first of its kind, we had no basis upon which to between lie- and truth-telling witnesses who wore niqabs, t(95) ϭ predict effect sizes for use in an a priori power analysis for Study 5.69, p Ͻ .001, d ϭ 0.58, 95% CI [0.36, 0.80] or hijabs, t(96) ϭ 1. We set a healthy sample size (i.e., 75 participants per veiling 8.98, p Ͻ .001, d ϭ 0.91, 95% CI [0.46, 1.37]. Participants could condition) and ceased data collection when our target was reached. not discriminate between lie- and truth-tellers who did not wear Due to the nature of our university’s participant pool (i.e., testing veils beyond chance levels, however, t(96) ϭ 0.45, p ϭ .652, d ϭ sessions were posted online at least one week in advance and 0.46, 95% CI [0.02, 0.07]. participants could modify appointments up until the beginning of Response bias. According to a Country ϫ Veiling Condition each session), our final sample size was slightly above what was ANOVA, participants’ biases (i.e., ␤) were affected by veiling specified. A post hoc power analysis of the discrimination find- ء 2 condition, F(2, 281) ϭ 5.03, p ϭ .007, ␩p ϭ .04 (see Table 1). Post ings, using G Power (Faul, Erdfelder, Lang, & Buchner, 2007), hoc tests indicated that participants who viewed witnesses in hijabs revealed that the study was adequately powered (power ϭ .97). By displayed a different pattern of response bias than those who saw using the effect size from the discrimination findings, we were able witnesses who did not veil, p ϭ .005, d ϭϪ0.44, 95% CI to estimate the required sample to produce statistical power at the [Ϫ0.72, Ϫ0.15]. Participants who viewed witnesses in niqabs did same level in Study 2 (i.e., N ϭ 290); we terminated data collec- not differ from those who saw witnesses in hijabs, p ϭ .335, d ϭ tion when it was reached. Thus, there is no reason to believe that

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 27 408 LEACH ET AL.

“p-hacking” was responsible for our significant lie detection re- lands—a country that had considered banning veils (e.g., Govern- sults. ment of the Netherlands, 2012)—might have been less positive Increases in lie detection accuracy associated with veiling might toward witnesses who wore niqabs, for example. Instead, partici- be attributed to the added emphasis on witnesses’ eyes. Partici- pants in the Netherlands, Canada and the United Kingdom viewed pants reported that they were no more likely to use the eye region veiled witnesses similarly. Judges and jurors always know whether to detect deceit when witnesses wore niqabs than when they did a witness is wearing a niqab while testifying and, presumably, not veil. Eye-tracking data suggest that, when forming social would exhibit the same tendencies as participants in our study. impressions, people spend more time looking at the eyes than any Indeed, meta-analyses have failed to find consistent differences in other feature (Janik, Wellends, Goldberg, & Dell’Osso, 1978). lie detection performance between students, community members, People’s eyes, and their perceived link to deception, might be so and justice officials (Aamodt & Custer, 2006; Bond & DePaulo, salient that highlighting them with a niqab was superfluous. In- 2006). deed, over 90% of the participants in our study reported using eye An additional limitation of this work is that we randomly contact as a cue to deceit whether the witnesses veiled or not. assigned our witnesses to lie and/or wear a veil. This practice was However, self-report should be treated with a degree of caution important from a scientific standpoint because it helped to ensure (e.g., Nisbett & Wilson, 1977). In our study, lie-tellers were more initial equivalence between the groups. Experimentally manipulat- likely to avert their gaze than truth-tellers; veils should have ing lying (vs. inducing volitional, naturalistic lies) should not have highlighted this difference. Improvements in lie detection perfor- significantly affected the results and is in keeping with previous mance suggest that participants might have attended to, or inter- research on lie detection (see Vrij, 2008 for a review). Witnesses preted, eye gaze information more accurately in the veiling con- thought that the study was involving, and they were motivated to ditions. be believed: the deception paradigm invoked experimental realism. Deception detection strategies were also affected by the amount However, because we randomly assigned witnesses to veiling of visual information that was available. Compared to the other condition, we might also have obscured natural differences be- conditions, witnesses in niqabs revealed significantly more verbal tween the groups (Ammar & Leach, 2013). For example, in Am- than nonverbal cues. Appropriately, participants were more likely mar and Leach’s (2013) study, the women who wore niqabs were to base their decisions on verbal cues than nonverbal cues when less likely to be native English speakers than women who did not viewing witnesses from this group. During several testing sessions, veil. Emerging work suggests that laypersons and police officers participants did not watch all of the videos (i.e., they turned away are not only less able to discriminate between lie- and truth-tellers from the screens and listened to the testimony). However, this who are speaking in a non-native language, but also view them practice only seemed to occur when witnesses wore niqabs. Future less positively than native speakers (Leach & Da Silva, 2013). It is research should examine the frequency of self-selected minimiza- unknown how natural variations in veiled witnesses’ language tion of information (e.g., using eye tracking). Establishing that proficiencies would have mitigated our findings. In the future, observers watched the witnesses would then allow researchers to researchers might wish to examine people’s assessments of actual explore the specific mechanisms underlying decision-making (e.g., niqab-wearers to address this issue. correlate deception cues with deception judgments using a lens The two studies reported here provide unique tests of the be- model analysis; see Hartwig & Bond, 2011). havioral assumptions underlying important courts decisions in the Despite not being explicitly discussed by the courts in the United States, United Kingdom, and Canada. The essence of these aforementioned cases, we considered whether response bias af- decisions is that women must remove their niqabs while testifying fected decisions related to veiled witnesses. The decisions of to ensure the fairness of court proceedings (e.g., The Queen v. judges and other members of the justice system are typically D(R), 2013). Although preliminary, in the sense that we have guided by principles related to fair treatment, such as those laid out reported only two empirical studies addressing these assumptions, in the Equal Treatment Bench Book in the United Kingdom the data consistently suggested that minimizing visual information (Judicial College, 2013). The same might not be true of jurors. actually improved participants’ lie detection performance. It is Tending to (dis)believe a veiled witness due to preexisting stereo- noteworthy that witnesses themselves believed that they would be types would severely undermine court proceedings. It was, thus, more accurately judged when wearing niqabs. Thus, seeing a encouraging that participants were not negatively biased against person’s entire face does not appear to be necessary for lie detec- witnesses who wore niqabs, even in the absence of explicit in- tion; banning the niqab because it interferes with one’s ability to struction. These findings replicated previous work, in which mock determine whether the speaker is lying or telling the truth is not jurors were similarly unaffected when a witness was described as supported by scientific evidence. In addition to the potential policy having worn a burqa (Maeder et al., 2012). implications concerning the wearing of a niqab or hijab on the We cannot completely discount the possibility that findings stand, the studies reinforce the value that behavioral science data were due to social desirability, however, because participants were have for informing judiciaries. not blind to veiling condition. Of course, if participants altered their responses systematically, that could not explain the above- chance discrimination between lie- and truth-tellers in the veiling References conditions (i.e., response bias and discrimination are independent; Aamodt, M. G., & Custer, H. (2006). Who can best catch a liar? A Green & Swets, 1966). Only response biases should have been meta-analysis of individual differences in detecting deception. Forensic affected. Yet, were our findings merely a reflection of socially Examiner, 15, 6–11. acceptable norms, then we might have expected differences in Akehurst, L., Kohnken, G., Vrij, A., & Bull, R. (1996). Lay persons’ response biases between the countries; participants in the Nether- and police officers’ beliefs regarding deceptive behavior. Applied

28 ■ Women in the Law ■ February 2017 VEILED WITNESSES 409

Cognitive Psychology, 10, 461–471. http://dx.doi.org/10.1002/ biomedical sciences. Behavior Research Methods, 39, 175–191. http:// (SICI)1099-0720(199612)10:6Ͻ461::AID-ACP413Ͼ3.0.CO;2-2 dx.doi.org/10.3758/BF03193146 Albrechtsen, J. S., Meissner, C. A., & Susa, K. J. (2009). Can intuition Fischer, P., Greitemeyer, T., & Kastenmuller, A. (2007). What do we think improve deception detection performance? Journal of Experimental about Muslims? The validity of Westerners’ implicit theories about the Social Psychology, 45, 1052–1055. http://dx.doi.org/10.1016/j.jesp.2009 associations between Muslims’ religiosity, religious identity, aggression .05.017 potential, and attitudes toward terrorism. Group Processes & Intergroup al-Ghazali, M. (2008). Lisat min al-islam [Not from Islam]. Cairo, Egypt: Relations, 10, 373–382. http://dx.doi.org/10.1177/1368430207078697 Dar al-shorouq. The Global Deception Research Team. (2006). A world of lies. Journal of Ambady, N. (2010). The perils of pondering: Intuition and thin slice Cross-Cultural Psychology, 37, 60–74. http://dx.doi.org/10.1177/ judgments. Psychological Inquiry, 21, 271–278. http://dx.doi.org/10 0022022105282295 .1080/1047840X.2010.524882 Government of the Netherlands. (2012, January 27). Government approves Ambady, N., Bernieri, F. J., & Richeson, J. A. (2000). Toward a histology ban on clothing that covers the face. Retrieved from http://www of social behavior: Judgmental accuracy from thin slices of the behav- .government.nl/news/2012/01/27/government-approves-ban-on- ioral stream. In M. Zanna (Ed.), Advances in experimental social psy- clothing-that-covers-the-face.html chology (Vol. 32, pp. 201–251). San Diego, CA: Academic Press. Green, D. M., & Swets, J. A. (1966). Signal detection theory and psycho- Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as physics. New York, NY: Wiley. predictors of interpersonal consequences: A meta-analysis. Psychologi- Hartwig, M., & Bond, C. F., Jr. (2011). Why do lie-catchers fail? A lens cal Bulletin, 111, 256–274. http://dx.doi.org/10.1037/0033-2909.111.2 model meta-analysis of human lie judgments. Psychological Bulletin, .256 137, 643–659. http://dx.doi.org/10.1037/a0023589 Ammar, N., & Leach, A.-M. (2013, November). Veiling in the courtroom: Hoodfar, H. (1997). The veil in their minds and on our heads: Veiling Muslim women’s credibility and perceptions. Paper presented at the practices and Muslim women. In L. Lowe & D. Lloyd (Eds.), The Qatar Foundation Annual Research Conference, Doha, Qatar. politics of culture in the shadow of capital (pp. 248–279). Durham, NC: Baron-Cohen, S., Wheelwright, S., & Jolliffe, T. (1997). Is there a “lan- Duke University Press. guage of the eyes”? Evidence from normal adults, adults with autism or Janik, S. W., Wellens, A. R., Goldberg, M. L., & Dell’Osso, L. F. (1978). Asperger syndrome. Visual Cognition, 4, 311–331. http://dx.doi.org/10 Eyes as the center of focus in the visual examination of human faces. .1080/713756761 Perceptual and Motor Skills, 47, 857–858. http://dx.doi.org/10.2466/ Behiery, V. (2013). Bans on Muslim facial veiling in Europe and Canada: pms.1978.47.3.857 A cultural history of vision perspective. Social Identities: Journal for the Jarvis, B. G. (2008). MediaLab (Version 2008.1.33e) [Computer software Study of Race. Nature and Culture, 19, 775–793. and manual]. New York, NY: Empirisoft Corporation. Bill 60: Charter Affirming the Values of State Secularism and Religious Judicial College. (2013). Equal treatment bench book. London, England: Neutrality and of Equality between Women and Men, and Providing a Judicial College. Framework for Accommodation Requests, Quebec National Assembly, Khiabany, G., & Williamson, M. (2008). Veiled bodies—Naked racism: 40th Legislature. (2013). Culture, politics and race in the sun. Race & Class, 50, 69–88. http:// Blais, C., Roy, C., Fiset, D., Arguin, M., & Gosselin, F. (2012). The eyes dx.doi.org/10.1177/0306396808096394 are not the window to basic emotions. Neuropsychologia, 50, 2830– Leach, A. M., & Da Silva, C. S. (2013). Language proficiency and police 2838. http://dx.doi.org/10.1016/j.neuropsychologia.2012.08.010 officers’ lie detection performance. Journal of Police and Criminal Bond, C. F., Jr., & Atoum, A. O. (2000). International deception. Person- Psychology, 28, 48–53. http://dx.doi.org/10.1007/s11896-012-9109-3 ality and Social Psychology Bulletin, 26, 385–395. http://dx.doi.org/10 Leal, S., & Vrij, A. (2008). Blinking during and after lying. Journal of .1177/0146167200265010 Nonverbal Behavior, 21, 87–102. Bond, C. F., Jr., & DePaulo, B. M. (2006). Accuracy of deception judg- Loi n° 2010–1192 du 11 octobre 2010 interdisant la dissimulation du ments. Personality and Social Psychology Review, 10, 214–234. http:// visage dans l’espace public [Act No 2010-1192 of 11 October 2010 dx.doi.org/10.1207/s15327957pspr1003_2 prohibiting the concealing of the face in public], J. O. 0237, October 12, Bond, C. F., Jr., Omar, A., Mahmoud, A., & Bonser, R. N. (1990). Lie 18344 (2010). detection across cultures. Journal of Nonverbal Behavior, 14, 189–204. Loi visant a` interdire le port de tout vêtement cachant totalement ou de http://dx.doi.org/10.1007/BF00996226 manière principale le visage, Moniteur, July 13 (2011). Burgoon, J. K., Blair, J. P., & Strom, R. E. (2008). Cognitive biases and Maeder, E. M., Dempsey, J., & Pozzulo, J. (2012). Behind the veil of juror nonverbal cue availability in detecting deception. Human Communica- decision-making: Testing the effect of Muslim veils and defendant race tion Research, 34, 572–599. http://dx.doi.org/10.1111/j.1468-2958.2008 in the courtroom. Criminal Justice and Behavior, 39, 666–678. http:// .00333.x dx.doi.org/10.1177/0093854812436478 Centre for Canadian Language Benchmarks. (2010). Canadian language Manson v. Brathwaite, 432 U.S. 98 (1977). benchmarks: English as a second language for adults. Retrieved on Meissner, C. A., & Kassin, S. M. (2002). “He’s guilty!”: Investigator bias August 25th, 2011, from http://www.cic.gc.ca/english/pdf/pub/ in judgments of truth and deception. Law and Human Behavior, 26, language-benchmarks.pdf 469–480. http://dx.doi.org/10.1023/A:1020278620751 Clarke, L. (2013). Women in niqab speak: A study of niqab in Canada. Mistry, H., Bhugra, D., Chaleby, K., Khan, F., & Sauer, J. (2009). Veiled com- Toronto, Ontario, Canada: Canadian Council of Muslim Women. munication: Is uncovering necessary for psychiatric assessment? Transcultural Datavyu Team. (2014). Datavyu: A video coding tool. New York, NY: Psychiatry, 46, 642–650. http://dx.doi.org/10.1177/1363461509351366 Databrary Project, New York University. Retrieved from http://datavyu Mok, T. (2013, September 12). “We don’t care what’s on your head”: .org Ontario hospital launches ad aimed at Quebec medical students, values DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, charter. The National Post. Retrieved from http://news.nationalpost.com/ K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin, 2013/09/12/we-dont-care-whats-on-your-head-toronto-area-hospital-ad- 129, 74–118. http://dx.doi.org/10.1037/0033-2909.129.1.74 aims-at-quebec-medical-students-values-charter/ ,Power 3: A Muhammad v. Enterprise Rent-A-Car, No. 06–41896-GC (31st D. MichءFaul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G flexible statistical power analysis program for the social, behavioral, and Oct. 11, 2006).

Effective Advocacy—Taking It to the Next Level ■ Schwartz ■ 29 410 LEACH ET AL.

Murphy, K. R., & Balzer, W. K. (1986). Systematic distortions in memory- The Queen v. D(R) [2013] EW Misc 13 (CC). based behavior ratings and performance evaluations: Consequences for Troje, N. F. (2002). Decomposing biological motion: A framework for rating accuracy. Journal of Applied Psychology, 71, 39–44. http://dx analysis and synthesis of human gait patterns. Journal of Vision, 2, .doi.org/10.1037/0021-9010.71.1.39 371–387. http://dx.doi.org/10.1167/2.5.2 Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Vakulenko, A. (2007). “Islamic headscarves” and the European Conven- Verbal reports on mental processes. Psychological Review, 84, 231–259. tion on Human Rights: An intersectional perspective. Social & Legal http://dx.doi.org/10.1037/0033-295X.84.3.231 Studies, 16, 183–199. http://dx.doi.org/10.1177/0964663907076527 N. S. v. Her Majesty the Queen, et al., No. 33989 (SCC 2012). Vrij, A. (2008). Detecting lies and deceit: Pitfalls and opportunities (2nd R. v. N. S., 670 (ONCA 2010). ed.). New York, NY: Wiley. Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive Vrij, A., & Winkel, F. W. (1991). Cultural patterns in Dutch and Surinam psychology: Undisclosed flexibility in data collection and analysis al- nonverbal behavior: An analysis of simulated police/citizen encounters. lows presenting anything as significant. Psychological Science, 22, Journal of Nonverbal Behavior, 15, 169–184. http://dx.doi.org/10.1007/ 1359–1366. http://dx.doi.org/10.1177/0956797611417632 BF01672219 Street, C. N. H., & Richardson, D. C. (2015, December). The focal account: Indirect lie detection need not access unconscious, implicit Wells, G. L., & Quinlivan, D. S. (2009). Suggestive eyewitness identifi- knowledge. Journal of Experimental Psychology: Applied, 21, 342–355. cation procedures and the Supreme Court’s reliability test in light of http://dx.doi.org/10.1037/xap0000058 eyewitness science: 30 years later. Law and Human Behavior, 33, 1–24. Syed, Z. (2010, June 21) Banning the niqab. Islamic Insights. Retrieved http://dx.doi.org/10.1007/s10979-008-9130-3 from http://www.islamicinsights.com Wixted, J., & Lee, K. (2014). Signal detection theory. http://dx.doi.org/10 Taylor, P., Larner, S., Conchie, S., & Van der Zee, S. (2014). Cross- .1002/9781118445112.stat06743 cultural deception detection. In P. A. Granhag, A. Vrij, & B. Verschuere Zuckerman, M., Koestner, R., & Colella, M. J. (1985). Learning to detect (Eds.), Deception detection: Current challenges and cognitive ap- deception from three communication channels. Journal of Nonverbal proaches (pp. 175–202). Chichester, England: Wiley Blackwell. Behavior, 9, 188–194. http://dx.doi.org/10.1007/BF01000739

Appendix Nonverbal and Verbal Cues

The nonverbal cues were eye contact, blinking, pupil dilation, Cues that were more difficult to quantify in that manner—vocal smiling, covering mouth and eyes, facial expressiveness, un- tension, coherence, vagueness, cooperativeness, nervousness, fa- friendly facial expressions, shifts in posture, self-manipulations cial expressiveness, generalizations, rate of speech, and amount of (e.g., self-touching or scratching), leg and foot movements, fidg- detail—were rated on a scale from 1 to 5. In addition, pupil eting, and use of hand gestures to illustrate speech. Vocal cues dilation was not coded by research assistants because it was not included stuttering, grammatical errors, repetitions of words or sufficiently visible in all videos. phrases, voice pitch, vocal tension, rate of speech, speech hesita- All of the nonverbal and verbal cues were presented to partic- tions, number of pauses, length of pauses, coherence of account, ipants as part of the Cue Use Measure. Participants indicated that length of answers, amount of detail, inclusion of unusual details, they had used the cue by selecting the box next to the word or spontaneous corrections, admitting lack of memory, inconsistent phrase. information, generalizations, vagueness, complaints, cooperativeness, and overall nervousness. Received March 16, 2015 When coding cues, research assistants counted the frequency of Revision received January 12, 2016 the majority of the nonverbal and verbal behaviors listed above. Accepted February 22, 2016 Ⅲ

30 ■ Women in the Law ■ February 2017