The influence of confession inconsistencies on juror decision-making

False confessions occur at a rate that contradicts the commonsense belief that only the guilty confess. Wrongful conviction statistics show that more than a quarter of people exonerated due to retrospective testing of DNA evidence were initially convicted based either solely or partly on their false confession to the crime. This has prompted a large body of research, which has contributed to understanding of how and why such false confessions occur, resulting in recommendations to reduce their instance. Recently, researchers have begun to investigate why juries accepted these false confessions in the first place, with case studies confirming that jurors are prone to accepting confession evidence that, in hindsight, should have been rejected. However, the picture is not entirely bleak, with other empirical evidence indicating that jurors will sometimes (albeit rarely) reject confession evidence that they view as potentially false. The basis of the variance in juror decision making on confession evidence is the focus of the current series of studies. In four chapters, we describe a series of mock-juror experiments investigating how jurors process confession evidence, and why they will sometimes accept a problematic, low quality confession as true, while at other times will reject the confession. Using a combination of typed confession statements, transcripts of police interview, and police summary statements, we aimed to identify the situational factors (confession attributes) that influence juror judgments of suspect guilt. Additionally, using existing scales (Need for Cognition, and Attitudes Toward Coerced Confessions) we tested the contribution of dispositional factors (juror attributes) in moderating juror decision making. We found that the confession attribute of consistency significantly influenced juror judgments of suspect guilt, but that this effect differed dependent on the type of inconsistency present in the confession (Chapter 2). Mock-jurors were not concerned about the suspect contradicting and then self-correcting their previous statements, but when the suspect made errors that could be proven factually incorrect with a secondary piece of evidence, judgments of suspect guilt were reduced. This particular finding suggests that jurors might more easily rationalize why a suspect might be telling a narrative in a way that requires backtracking and correction (e.g. deliberate mistruth, confusion due to intoxication), than why a suspect would get key facts of the crime entirely wrong. However, further testing revealed that not all factual errors were treated equally (Chapter 3). When factual errors were parsed out into errors that appeared to be amplifying or downplaying the crime severity, we found differential effects of directionality on judgments of guilt that were not aligned with how we had hypothesized jurors would process such inconsistencies. For example, we believed that if the suspect confessed to a crime greater than the facts would imply, participants might infer that he is making himself look worse, and wonder why anyone would confess to a greater crime than they actually committed, thus discounting the confession. On the other hand, if the suspect confessed to a lesser version of the confirmed crime, he might be seen as attempting to make himself look better, or less culpable, which has a clear ulterior motive, and would result in the juror upholding the confession. Contrary to these hypotheses, no discounting was applied to confessions where the suspect appeared to be making his case worse by admitting to a more severe version of the crime, in comparison to a confession in which the suspect confessed fully and without error. While this finding speaks to the overall believability of confession evidence, and that jurors will simply accept any confession as true, results of the better (i.e. decreased crime severity) confession provide interesting data for consideration. When factual errors acted to decrease crime severity, (thereby making the suspect look better) mock-jurors judged the suspect as more guilty of the crime than the suspect who admitted fully and accurately to the crime, effectively adding a punitive cost to perceived downplaying of the crime. Overall, we found that mock-jurors did notice confession inconsistencies, and that some types of inconsistencies, in some circumstances, would influence judgments of suspect guilt. While chapters 2 and 3 focused on situational factors that might account for the variance in juror decision-making when presented with confession evidence, the papers described in chapters 4 and 5 were concerned with identifying dispositional factors that might interact with confession attributes to influence judgments of suspect guilt, with mixed success. Need for cognition (NC) has been identified as a possible moderator in juror decision making, as it is concerned with inherent levels of motivation to engage in cognitively challenging tasks (Chapter 4). One reason posited for why jurors sometimes accept confession evidence that should have been rejected due to poor quality, is that jurors do not adequately scrutinize confession evidence, as they implicitly (and logically) believe it to be truthfully given. However, the naturally elevated motivation of high NC jurors should translate to an increased likelihood of engaging with the confession evidence more deeply and analytically, moderating any effect of confession inconsistencies on judgments of guilt. While previous research would generally agree that NC moderates juror decisions in some capacity, the current study and other studies using confession stimuli have not found this to be true. Across four experiments, we found no effect of need for cognition (measured on the 18-question Need for Cognition scale) on judgments of suspect guilt. We posit a number of reasons for this null finding, including the possibility that the motivation component of need for cognition does not adequately control for engagement levels in a trial where confession evidence is present. However, while need for cognition failed as a dispositional factor contributing to juror variance in our study, we had considerable success with another juror attribute; that of attitudes toward coerced confessions (Chapter 5). Using the previously untested Attitudes Toward Coerced Confessions scale (ATCC) we measured underlying support for coercive interrogation tactics and the belief that false confessions can be coerced from an innocent person. Findings showed the ATCC to be a valid measure of its two stated constructs (i.e. support for coercive interrogation, and belief in coerced confessions), and was reliably replicated in an experiment where coercion was manipulated within the confession. We found that the subscales of the ATCC could accurately predict which participants would, or would not, reduce their perception of suspect guilt when inconsistencies or coercion were present in the confession. In summary, the present series of experiments adds to the literature on juror decision making by addressing the mechanisms underpinning the ways in which jurors process confession evidence. Our findings show that while jurors can discern confession inconsistencies, recognition of those inconsistencies only result in reduced judgments of guilt in certain circumstances. While need for cognition failed to account for any variance in juror decision making, one dispositional attribute that moderated judgments of suspect guilt was identified. The ATCC scale was able to successfully predict which individuals would ignore the coercive elements of a confession when deciding if a confession was reflective of true guilt, as well as those individuals who would be concerned by the inconsistencies in a confession and reduce their judgment of guilt accordingly.