Skip navigation
The Habeas Citebook: Prosecutorial Misconduct - Header
× You have 2 more free articles available this month. Subscribe today.

Research Paper Reveals Laypeople Have Insufficient Understanding of False Confessions by Examining Prior Research Based on Surveys and Mock Juries

by David M. Reutter

Research via use of surveys and mock juries leave unexplained the “possible explanations for the often-found lack of adjustment for interrogation tactics reflected in verdicts among mock jurors.” This article summarizes a report that peered into studies by third parties that conducted experiments using these two research models.

The 44-page report, titled Do Jurors Understand the Causes of False Confessions, and Do They Adjust Their Perceptions of Suspects’ Confessions Appropriately,was written by Dr. Deborah Davis, Professor of Psychology at the University of Nevada, and Dr. Richard A. Leo, the Hamill Family Professor of Law and Psychology at the University of San Francisco (together, “Authors”).

The report begins with a case study of the September 26, 1983, rape and murder of 11-year-old Sabrina Buie. Based upon a report of a rumor by a 17-year-old high school student, police focused on two brothers. Henry McCollum, 19, and Leon Brown, 15, had IQs of 51 and 49, respectively. They were threatened with the death penalty and accused of the rape and murder during the overnight interrogation, which lasted about five hours.

Police persuaded the illiterate brothers to sign the written statements prepared by the police with promises that they could go home if they did so. No other evidence linked them to the crime. Both were convicted of capital murder and sentenced to death. Their convictions were reversed on appeal, but the brothers were again convicted. McCollum once again was sentenced to death while Brown received a life sentence. They were subsequently exonerated by DNA evidence, their convictions were vacated, they were released after serving 31 years, and North Carolina Gov. Patrick McCrory pardoned them in 2015.

“As illustrated by the cases of Henry McCollum and Leon Brown, those in the legal system who must judge the veracity of a confession commonly lack the requisite knowledge to recognize when a confession is false,” wrote the Authors. “Studies of proven false confessors have shown that when their cases go to trial, between 73 and 81 per cent are convicted.”

“Neither law enforcement, prosecutors nor jurors were able to correctly judge the validity of their confessions. Such errors are not surprising given the very large range of perceiver knowledge and contextual information necessary to evaluate the potential for false confession. These include chronic and acute vulnerabilities of the suspect, the nature of the accusations and of the interrogation, the reasons why suspects falsely confess, how interrogation tactics persuade and more.”

“Interrogation scholars have accumulated a wealth of data testing the impact of personal and situational factors affecting the likelihood of a false confession,” continued the Authors in their Introduction. “Yet, when judges have excluded expert testimony on interrogation and confessions, a common justification has been that the information an expert could provide is not beyond common knowledge and therefore not helpful to the jury.”

That claim was addressed by the Authors “through a review of scientific studies assessing lay beliefs concerning police interrogation and confession, and a comparison of the content of these lay beliefs to the findings of relevant interrogation science,” said the Authors. “Two predominant methods have been used to study lay understanding of interrogations and confession: (i) surveys of beliefs regarding interrogation and confession; and (ii) mock jury studies.”

“The surveys primarily assess core beliefs about interrogation and confession, largely free of the context of case facts, the broader context of interrogation or the combined impact of multiple tactics and vulnerabilities. These beliefs can be regarded as the expectations with which observers might approach judgment of interrogations and confessions in case contexts.” In “mock jury studies, participants are presented with case summaries in which characteristics of the defendant or the interrogation are varied and the mock juror’s reactions are assessed. The intent is to assess whether the mock jurors react to the variations in the way interrogation science suggests that they should.”

Surveys generally ask “participants to rate the likelihood that police would use various tactics. This method is, of course, reactive in the sense that participants may have had no thoughts about these tactics until these are presented in the survey and the participants must provide a response. Therefore, we cannot know from these surveys whether there is general awareness of the use of various tactics. However, the surveys do, at a minimum, indicate what laypersons view as the most likely used among the tactics presented.”

The Authors detailed the findings of survey researchers in nine different areas of interest. The first concerned the tactics interrogators use. Most of those surveyed in one study reflected that it was understood that police use tactics such as “subjecting the suspect to lengthy interrogations of several hours or more; promising more lenient treatment contingent on a confession; pretending to befriend the suspect; threatening with more severe consequences in the absence of a confession; repeated statements of confidence in the suspect’s guilt; depriving suspects of social contact and support; and falsely claiming evidence against the suspect. Deprivation of food, water or sleep were rated close to the midpoint of the scale. When asked specifically about whether the police lie about evidence, most agreed that they do (mean rating of roughly 5 on a 7-point scale) though most viewed this as unacceptable (2 on a 7-point scale).”

In another study, “participants believed that the police use threats or actual physical harm (14 per cent) and relatively few believed the police use rapport-building (56 per cent) or false evidence (42 per cent). The most commonly believed tactics were confrontation with evidence of guilt (83 per cent), bluffs about evidence (78 per cent), promises of leniency (74 per cent), and rejecting suspect denials (64 per cent). The authors also found that participants believed interrogations generally last more than eight hours.”

The perceptions of what is permissible in an interrogation is far different than what occurs when interrogators are giving a suspect the “Third Degree.” The Authors found that studies “indicated that roughly 60 per cent or more of participants believe that the police are not allowed to lie to suspects. Yet, despite this, courts in the U.S. have upheld police tactics involving lying or deception, including about evidence. Some interrogation manuals used in the U.S. and some other countries encourage police to misrepresent evidence as an effective way to elicit confessions, and 92 per cent of the police report that they sometimes lie about evidence.”

One notable study focused on misconceptions among participants and found that “56 per cent believed that the police cannot cut off suspect denials, 45 per cent believed that the police cannot downplay the significance of a crime, 59 per cent believed that the police cannot use rude or insulting remarks and 43 per cent believed that it is permissible to threaten harsher punishment if the suspect does not confess. Otherwise, participants were generally informed concerning the permissibility of such tactics as deprivation of food, water or sleep, minimization, implications of leniency and others. Highly educated and white respondents were more likely to be correct regarding these issues of legality. Eighty per cent agreed that expert testimony would be helpful in cases involving disputed confessions.”

Studies have found that survey participants believe psychology tactics are acceptable, but they believe physical tactics are less appropriate. “[T]here was no main effect of crime severity. There was, however, an interaction of tactics and crime severity: such that severe tactics were rated as more appropriate for severe crimes, but crime severity did not affect the perceived appropriateness of psychological tactics. Further, the tactics also interacted with evidence strength. When evidence was weak, physical tactics were rated as less appropriate than when evidence was strong. The opposite was found for psychological tactics.”

One of the studies reviewed by the Authors found tactics “rated by the fewest participants as highly coercive were rapport-building (34 per cent), confronting suspects with true evidence of guilt (47 per cent) and rejecting suspect denials (53 per cent). All others were rated as highly coercive by 64–69 per cent of participants. These included bluffs about evidence, use of false evidence, promises of leniency and threats and use of physical harm. Participants familiar with false confession cases rated only two tactics as more coercive than unfamiliar participants—promises of leniency and threatening or using physical harm.”

As to the question of whether false confessions occur and why, the Authors reported that a “substantial minority of 32 per cent held” the “conviction that false confessions simply do not occur.” The study “further found that 46 per cent believed physical torture was the most likely path to false confession, followed by confession to receive a lesser charge (18 per cent), confession to cover for someone else (18 per cent), confusion and a belief that he or she might actually be guilty (9 per cent), overall stress of the interrogation (7 per cent), and need for notoriety (1 per cent).”

In contrast, another study “found that only 6 per cent believed that innocent suspects would never confess, whereas most believed they would do so only after strenuous pressure.”

“Participants also responded to the statements ‘Some suspects may be too stressed to offer a reliable confession’ and ‘A suspect might confess to a crime just to end the police interrogation’. These ratings were all approximately 4.5 on a 6-point scale, indicating greater agreement that false confessions can occur than that they cannot.”

Participants also believed that a confession is highly indicative of guilt. One study “found that 64 per cent of participants agreed that a confession was a strong indicator of guilt, and that 50 per cent of them agreed that if a person confesses, they are probably guilty,” the Authors wrote. “Interestingly, only 7 per cent indicated that a person who signed a written confession in interrogation is definitely guilty (and 59 per cent believed that he is probably guilty).”

Next, the Authors examined the perceptions of tactics that lead to false confession. “Tactics rated as likely to result in a false confession by the most participants were threatening or using physical harm (68 per cent), confronting suspects with false evidence (63 per cent), bluffing about evidence (58 per cent), promises of leniency (54 per cent), and rejecting suspect denials (53 per cent): and mean ratings of likelihood were above the midpoint for all of these. Confrontation with true evidence of guilt (21 per cent) and rapport-building (23 per cent) were viewed as likely cause false confession only by a minority, and mean ratings of likelihood for these were below the midpoint. Repeated accusations and cutting off denials were least commonly considered likely to result in a false confession (16–19 per cent).”

One study found that participants believed they would not falsely confess. “Nevertheless, when subjected to an actual interrogation themselves, [ ] 46 innocent subjects were twice as likely to confess when interrogation techniques implied consequences than when they did not (42 per cent versus 21 per cent).”

Finally, the perceptions of surveyors as to when a confession should be suppressed was examined. “Only failure to read Miranda rights, denial of a request for an attorney, denial of food or water, and physical assault were viewed as grounds for suppression by 64–65 per cent of participants. Threatening and intimidation were viewed as such by a small majority (54 per cent). But none of the other previously studied seven tactics (see above) was viewed as grounds for suppression by a majority (implied and explicit promises of leniency were viewed as such by roughly only a third of participants).”

“Some of the most well-established findings among interrogation studies concern the enhanced susceptibility of several vulnerable populations to false confessions, including the young, those with low IQ or a mental disability or mental illness, and many others. Other research has identified acute vulnerabilities posed by the physical or mental state of the suspect at the time of interrogation. Still others have pointed to the importance of minority status or type of accusation. Accordingly, many surveys have assessed lay awareness of these vulnerabilities, though the latter two remain unaddressed in surveys.”

Surveys of police found they “generally believed that children can be treated similarly to adults in interrogations and reported that they used similar interrogation tactics for both. Interestingly, more highly ranked or experienced officers tended to believe more strongly that suspects understand their rights and the intent of interrogation, and less strongly believe that interrogation tactics can lead to false memories or false confessions. Overall, the police believed that they elicited false confessions from about 10 per cent of suspects.”

Surveys of laypersons “found that many respondents were either uncertain or denied that children (57 per cent), the mentally impaired (46 per cent) or those who believe that they have faulty memories (74 per cent) are more vulnerable to interrogation-induced false confessions.” Researchers asking participants to “rate the extent to which various vulnerabilities would contribute to a person falsely confessing, and found that mental illness was viewed as most likely to do so (5.5 on a 7-point scale), followed by being under 10 years of age (4.6), being suggestible or overly trusting (4.5), possessing a low IQ (4.5), being a teenager (4.0), and having a poor memory (3.5).”

The Authors then examined conclusions to be gathered from mock jury studies. Study “results of available research are generally consistent with the idea that jurors will adjust verdicts based on interrogation length. More research is needed, however, as most studies confounded length with other interrogation tactics, and one of the two that did not, did not examine verdicts.”

Jurors viewing a high-pressure interrogation “viewed the confession elicited in the high-pressure interrogation as less voluntary and reported that it had less influence on their decisions, it nevertheless did affect their verdicts: defendants who confessed were regarded as more guilty regardless of interrogation pressure.” Researchers found that whether a confession was low-, medium-, or high-pressure; it had no impact on the verdict.”

When it came to false evidence ploys (“FEPs”), the verdicts were not studied, but participants were asked to rate “deceptiveness, coerciveness and justification. Testimonial FEPs were viewed as more deceptive and coercive than demeanor FEPs. No other differences were significant. However, participants who had been told that FEPs lead to true confessions viewed them as more justified.” A more recent study “found discounting of suspect confessions due to an informant incentivized to report an alleged defendant confession, though several prior studies had not found this effect.”

Different populations’ vulnerability to false confessions was then scrutinized. Researchers “found no effect of an interrogated suspect’s age (11 versus 14 years) on ratings of guilt (though a ceiling effect is possible in that 76 per cent judged the suspect guilty overall).” The presence of an adult during the interrogation increased consciousness conviction rates. “This finding might seem to indicate that the suspect was seen as less vulnerable to coercion when the adult was present, and accordingly that the confession was seen as more voluntary. However, though voluntariness was not measured, the authors found that defendants were rated as more vulnerable with the adult present.”

Mental illness has been proven to play a role in vulnerability to false confessions. Researchers “found that that jurors discounted the confession if the suspect confessed due to concerns about a medical disorder, but not if he did so due to the pressures of the interrogation or owing to a mental disorder.”

Few studies have looked to the vulnerability of the intellectually disabled to giving a false confession. Most studies were conducted in juvenile defendants, creating double vulnerabilities. “For the intellectually disabled defendant, ratings of degree of guilt were lower (but not dichotomous verdicts), the confession was viewed as less voluntary and the police were perceived as more coercive. Jurors felt they were less influenced by the disabled defendant’s confession, but yet did not rate the truthfulness of that confession any less.”

Temporary incapacities such as intoxicating and sleep deprivation were also reviewed. Little to no research has addressed juror reactions to these variables; nevertheless, the Authors found at least one study in each area to review. “[S]ober and intoxicated defendants who confessed were convicted at equal rates (47.6–47.9 per cent). Without a confession, intoxicated defendants were more likely to be convicted (44 per cent versus 34 per cent). Once again, although survey responses and the responses of mock jurors indicated that laypersons are aware of the impairment caused by intoxication, their verdicts did not reflect any adjustment for intoxication.” A recent study “found that recent sleep deprivation was linked to perceptions of the evidence strength of the confession and its likely validity and voluntariness (but see weaker results in the first study).”

Surveys have not addressed race as a factor in false confessions. “The presence of a confession increased the perception of guilt only when the crime was stereotypic for the defendant’s ethnicity; and did so regardless of the level of interrogation pressure. Even with a low-pressure interrogation, a confession did not increase perceived guilt if the crime was counter-stereotypic,” reported the Authors.

When looking at the characteristics of a confession, “the results of studies examining effects of the inconsistency between the content of a confession and the actual case facts indicate that laypersons are indeed sensitive to this issue. Notably, they also appear to adjust verdicts according to whether case facts were first raised by the interrogator or suspect,” the Authors wrote. “To the extent that jurors can be made sufficiently aware of who first introduced case facts, it appears they will be willing to adjust their perceptions of the import of the content of the confession accordingly.”

“At this point, the survey studies indicate lay appreciation of the potential for interrogations to be coercive and to elicit false confessions. Though these outcomes may be underestimated for some issues and overestimated for others, overall, laypersons recognize the potential impact for many specific interrogation factors and personal vulnerabilities,” concluded the Authors. “Moreover, mock jury studies reflect this knowledge in part. Nevertheless, it seems that research addressing juror knowledge or use of knowledge regarding interrogation and false confessions is at a somewhat unsatisfying point.” Thus, they called for more research on the issue.  


Source: Davis, Deborah and Leo, Richard A., Do Jurors Understand the Causes of False Confession, and Do They Adjust Their Perceptions of Suspects’ Confessions Appropriately? ( 2023). Forthcoming in Gavin Oxburgh, Trond Myklebust, Mark Fallon and Maria Hartwig, Eds. (2023). Interviewing and Interrogations: A Review of Research and Practice since World War II (Torkel Opsahl Academic EPublisher), Available at SSRN:

As a digital subscriber to Criminal Legal News, you can access full text and downloads for this and other premium content.

Subscribe today

Already a subscriber? Login



PLN Subscribe Now Ad
CLN Subscribe Now Ad
The Habeas Citebook Ineffective Counsel Side