Let's find the evidence: an analogue study of confirmation bias in criminal investigations

People involved in criminal proceedings (e.g. police officers, district attorneys, judges, and jury members) may run the risk of developing confirmation bias, or tunnel vision. That is, these parties may readily become convinced that the suspect is guilty, and may then no longer be open to alternative scenarios in which the suspect is actually innocent. This may be reflected in a preference for guilt‐confirming investigation endeavours, as opposed to investigations that are aimed at confirming, or even excluding, alternative scenarios. In three studies, participants read a case file, and were subsequently instructed to select additional police investigations. Some of these additional endeavours were guilt‐confirming (i.e. incriminating), whereas others were disconfirming (i.e. exonerating). Results suggest that additional investigation search was guided by an initial assessment of the suspect's guilt (Study 1). Furthermore, participants' tendency to select incriminating investigations increased with increased crime severity, and with the strength of the evidence present in the case file. Finally, the selection of incriminating investigations was associated with conviction rates (Study 3). However, in general, participants did not favour incriminating endeavours. That is, in the three studies, the percentages of selected incriminating endeavours did hardly or not exceed 50%. Copyright © 2010 John Wiley & Sons, Ltd.     

Second-Guessing the Jury: Stereotypic and Hindsight Biases in Perceptions of Court Cases1

This research examined the effects of stereotypic beliefs and hindsight biases on perceptions of court cases. Subjects read evidentiary material pertaining to a criminal trial in which the defendant either was a stereotyped offender or was not. Additionally, some subjects were given outcome information about the verdict attained in the trial; half of these subjects were told that the defendant had been found guilty, and the other half were told that he had been found not guilty. The remainder were not given any outcome information. Subjects were than asked to predict the likely outcome of a trial based on the presented evidence. Typical hindsight bias effects were expected and obtained for nonstereotyped offenders; subjects considering these cases viewed the evidence as less incriminating when they were told the defendant had been found not guilty, and they found it to be more incriminating when they were told the defendant had been found guilty, when compared to the no-outcome-information group. However, no hindsight biases were evident in judgments of cases involving stereotyped defendants, who were seen as relatively more likely to be guilty regardless of the nature of outcome information presented. Particularly striking was the lack of impact of the “not guilty” outcome information on perceptions of the guilt of stereotyped defendants. These findings suggest that strong expectations held in foresight may not be amenable to modification in hindsight.     

Judicial decision making: order of evidence presentation and availability of background information

An experiment was conducted to investigate both the effect of the order of presentation of defence and prosecution evidence and the prior availability of background information on assessment of guilt. Subjects were required to judge the defendant's probability of guilt either after each witness statement (step-by-step) or after having read all witness statements (end-of-sequence). In the step-by-step mode, an order effect was observed with later evidence exerting a greater impact on the subjects' judgment. This recency effect probably occurred because subjects used an anchoring-and-adjustment process: each new piece of evidence was averaged with an anchor judgment reflecting the overall assessment of previous items. In the end-of-sequence mode, on the other hand, the order effect depended on the background condition: if background information was provided a recency effect occurred, but if no background information was available a primacy effect was evident. This result might be explained by assuming that subjects tried to integrate witness information into a coherent cognitive pattern. As the judgment is memory-based in the end-of-sequence condition, recent information will be more available than earlier items. However, when no background information was presented, the first evidence items had to be processed at a more semantic, deeper level, resulting in a primacy effect that apparently outweighed the recency effect. © 1998 John Wiley & Sons, Ltd.     

Sexual experience and recall of sexual vs. nonsexual information

The effects of both sex guilt and sexual experience on the ability to remember sexual and nonsexual information were examined Male and female subjects were asked to read and then recall facts from vignettes that concerned sexual and nonsexual dilemmas As expected, the sexually inexperienced subjects made more mistakes than did the experienced subjects when recalling sexual information There were no differences between the groups in recall of nonsexual information, and there were no differences associated with the sex guilt variable In addition, one subgroup of the sample was singled out as being particularly interesting from an applied perspective Previous research has identified high-guilt, high-experience women as being an at-risk group for unwanted pregnancy In fact, these women did have difficulty in remembering the sexual information Moreover, there was a clear bias in their recall pattern toward remembering information in favor of being sexually active Results are presented in terms of the relationship between sexual experience and sexual schemata, and between sex guilt and emotionality     

Guilt assessment after retracted voluntary and coerced-compliant confessions in combination with exculpatory or ambiguous evidence

We investigated how voluntary confessions, coerced-compliant confessions, and no-confessions influenced guilt assessments in combination with other exculpatory or ambiguous evidence. In three experiments (total N = 808), participants studied case information and provided guilt assessments. As expected, in Experiment 1 and 2a, (i) voluntary confessions to protect a family member elicited stronger guilt attributions than no-confessions and (ii) ambiguous evidence led to stronger guilt attributions than exculpatory evidence. In Experiment 2b, voluntary confessions to protect a group-member (but not to protect a family-member) elicited stronger guilt attributions than no-confessions. Exculpatory eyewitness evidence elicited stronger guilt attributions than exculpatory DNA evidence and participants assigned more weight to exculpatory DNA than eyewitness evidence. Participants were able to discount coerced-compliant confessions when they received information about the interrogations (Experiments 2a/b), but did not consistently consider risk factors for (voluntary) false confessions outside the interrogation room when assessing guilt.     

How a “Reasonable Doubt” Instruction Affects Decisions of Guilt

Jurors are instructed to render a guilty verdict if they feel the defendant is guilty beyond a reasonable doubt. The jury is often told that this does not mean an absolute certainty of guilt and that even if it were possible to imagine a scenario in which the defendant is innocent, a guilty verdict may still be appropriate. Here, participants read a case summary. They were either told to say that the defendant was guilty if they believed in guilt beyond a reasonable doubt or were given more detailed instruction stressing that they did not have to be absolutely certain of guilt to give a guilty verdict. In Experiment 1, participants provided “think-aloud” protocols. Content analysis revealed that those who were given this instruction often used the phrase reasonable doubt to justify their guilty verdicts by saying that although they were not certain of the defendant's guilt, their belief exceeded the reasonable doubt threshold. None of the participants in the control group did this. Experiment 2 was designed to test if the instruction affected belief in guilt and the reasonable doubt threshold quantitatively. The instruction affected both people's belief in guilt and the threshold that they used to define reasonable doubt. The implied values for reasonable doubt were 63% for those who received the instruction and 77% for the control group. Implications for jury decision making are discussed.     

A study of the comparative structure of guilt and shame in a Chinese society.

On the basis of previous theoretical and empirical analyses of the comparative structures of guilt and shame, the authors hypothesized that antecedent condition (personal inadequacy vs. moral norm violation), audience presence, and personal responsibility attribution would distinguish shame from guilt. Although the subject population was Hong Kong Chinese, evidence from previous studies suggests that the comparative structures of guilt and shame are quite similar across cultures. The subjects were asked to recall either a guilt or a shame incident, and their responses were then coded into the predictor variables. The results of the study indicated that guilt was most likely to emerge when individuals had violated a moral norm and held themselves responsible for their conduct. In contrast, shame emerged more frequently when subjects felt personally inadequate than when they had violated moral norms. Moreover, when a guilt incident was reported, and audience was rarely mentioned, whereas subjects who reported a shame incident would generally feel personally responsible and often mentioned being looked at or evaluated. However, neither personal responsibility nor the presence of an audience seemed to be essential for a person to experience shame.     

Race,Rape, and Bias: Distortion of Prior Odds and Meaning Changes

Two experiments were conducted to investigate how racial bias affects juror decision making. Three sources of bias were studied: (1) prior probabilities of guilt, (2) distortion of the meaning of evidence, and (3) differential weighting of information. A paired comparison technique employed in the first study revealed that pretrial probabilities of guilt were greater when the victim was White than when she was Black. In the second experiment, a different group of subjects viewed one of four videotaped simulated rape trials in which seven segments of testimony had been previously rated as pro-prosecution, pro-defense, or neutral. During the trial, subjects rated each segment on three different scales: prosecution, defense, and degree of defendant guilt. Results indicated that neutral evidence was seen as more favorable to prosecution for a White victim compared to a Black victim. Evidence which favored either prosecution or defense was not distorted. Regression analyses revealed a positive relationship between estimates of guilt and distortion of evidence. The weight or importance of the evidence did not vary as a function of victim or defendant race. Years of recommended imprisonment indicated greater severity toward the Black assailant of a White woman. The results suggest that bias in favor of White victims occurs both in the assessment of pretrial probabilities and perception of evidence.     

Effects of pre‐trial publicity and jury deliberation on juror bias and source memory errors

We examined the effects of exposure to pre‐trial publicity (PTP) and jury deliberation on juror memory and decision making. Mock jurors either read news articles containing negative PTP or articles unrelated to the trial. They later viewed a videotaped murder trial, after which they either made collaborative group decisions about guilt or individual decisions. Finally, all participants independently attributed specific information as having been presented during the trial or in the news articles. Exposure to PTP significantly affected guilty verdicts, sentence length, perceptions of defendant credibility, and misattributions of PTP as having been presented as trial evidence. Jury deliberation had significant effects on jury verdicts, perceptions of defendant credibility, source memory for trial items, and confidence in source memory judgements, but did not affect sentences or critical source memory errors. Copyright © 2006 John Wiley & Sons, Ltd.     

Effects of inadmissible Evidence on the Decisions of Simulated Jurors: A Moral Dilemma

In a 2 × 3 design, simulated jurors received either weak or strong evidence against a defendant in a murder case. Within each evidence condition, subjects were given either additional evidence ruled admissible, additional evidence ruled inadmissible, or no additional evidence. Results indicated that ( a ) jurors were biased by inadmissible evidcnce in the weak-evidence but not in the strong-evidence condition, ( b ) strong evidence resulted in more guilty verdicts than did weak evidence, ( c ) confidence in verdict was influenced by strength of evidence, and ( d ) confidence of guilt was positively correlated with severity of punishment. Implications of the results are discussed within the context of the judicial process.     

Biases In Trials Involving Defendants Charged With Multiple Offenses1

Under federal and state laws, a defendant who has been charged with more than one offense can be tried for all the offenses in a single “joined” trial. It was predicted that the probabillty a defendant would be convicted would increase as a function of the number of joined offenses. Legal theories, research on memory, and social psychological models of information integration and attribution led to three hypotheses as to why this bias might occur: (1) confusion of evidence, (2) accumulation of evidence, and (3) inference of a criminal disposition. Subjects read and judged written trial summaries presented as joined or single trials. In Study 1, joinder resulted in higher rates of conviction and in confusion of evidence. In Study 2, the conviction results were replicated, and subjects judging joined trials also rated the evidence as more incriminating and made negative attributions about the defendant. These ratings were strongly related to judgments of guilt. A sequential judgment process was also found to affect jurors' judgments.     

Do apes know that they could be wrong?

When confronted with uncertain or incomplete information in decision-making situations, monkeys and apes opt for either escaping the situation or seeking additional information. These responses have been interpreted as evidence of metacognitive abilities. However, this interpretation has been challenged. On the one hand, studies using the information-seeking paradigm have been criticized because subjects may simply engage in a search for information routine (e.g., search until spot the reward) without any metacognitive involvement. On the other hand, studies using the escape response paradigm have been criticized because subjects may not recognize their own state of uncertainty but have learned to use the escape response in the presence of certain stimuli configurations that create uncertainty. The current study attempted to address these two criticisms by presenting great apes (seven gorillas, eight chimpanzees, four bonobos, seven orangutans) with a seeking information task whose basic procedure consisted of presenting two hollow tubes, baiting one of them and letting subjects choose. Conditions varied depending on whether subjects had visual access to the baiting, the cost associated with seeking information, the time interval between baiting and choosing, the food quality and the additional information offered regarding the food’s location. Although subjects showed a high retrieval accuracy when they had witnessed the baiting, they were more likely to check inside the tube before choosing when high stakes were involved (Experiment 3) or after a longer period of time had elapsed between the baiting and the retrieval of the reward (Experiment 2). In contrast, providing subjects with indirect auditory information about the food’s location or increasing the cost of checking reduced checking before choosing (Experiment 1). Taken together, these findings suggest that subjects knew that they could be wrong when choosing.     

Evaluating sex discrimination claims: the mediating role of attributions.

The role of attributions in judgments of sex discrimination was examined in 2 laboratory experiments. In Study 1, participants read 1 of 12 brief scenarios in which limited information about the strength of evidence against a fictitious corporation and occupational gender stereotype were manipulated. Results suggested that attributions mediated the relationships between participants' gender, strength of evidence, and discrimination judgments. In Study 2, participants were provided with 1 of 3 detailed, typewritten summaries of evidence presented in a sex discrimination trial. Results indicated that jurors' gender was again significantly related to attributions and to sex discrimination judgments even in the face of substantial objective information related to the case. The variance in observers' judgments associated with gender, however, appeared to be greatest when information about the organization's guilt or innocence was equivocal.     

Juror decision making: A case of attitude change mediated by authoritarianism

For a jury to reach a unanimous decision, certain individuals must change their attitudes with regard to the defendant's guilt during deliberations. Because these changers are the key to the group decision-making process, they were carefully scrutinized to ascertain demographic or personality characteristics which might be mediating their behavior. In three experiments, two using college students as subjects and one using Philadelphia Court of Common Pleas jury pool members, it was found that authoritarians changed their attitude with regard to the defendant's guilt more than equalitarians. Further, most of these “changers” were aware they had changed their attitude. There was no generalizable evidence for the proposition that authoritarians are more likely to favor a guilty verdict.     

The effectiveness of attentional diversion and situation redefinition for reducing stress due to a nonambiguous Threat

After receiving a sample shock, subjects in a Threat Condition were told that they would receive additional painful shocks while subjects in a Nonthreat Condition were not threatened with additional shocks. Subjects in an Attentional Diversion Condition were then instructed to read and think about an amusing story, subjects in a Situation Redefinition Condition were instructed to write down reasons why they should not be afraid in this situation, and subjects in a Control Condition were not given any instructions for coping with stress. (Coping Conditions were factorially crossed with Threat Conditions and 192 subjects were employed.) Physiological measures (pulse rate, finger pulse volume, skin resistance) indicated that unlike the results of previous research, attentional diversion was effective in reducing stress but situation redefinition was not. Discussion was focused on the parameters which determine whether or not a coping strategy will be effective.     

Negative and positive pretrial publicity affect juror memory and decision making

The experiment examined the effects of exposure to pretrial publicity (PTP) and delay on juror memory and decision-making. Mock jurors read news articles containing negative PTP, positive PTP, or unrelated articles. Five days later, they viewed a videotaped murder trial, after which they made decisions about guilt. Finally, all participants independently attributed specific information as having been presented during the trial or in the news articles. Half of the jurors rendered their verdicts and completed the source-memory test immediately after the trial, while the other half did so after a 2-day delay. Exposure to PTP significantly affected guilty verdicts, perceptions of defendant credibility, juror ratings of the prosecuting and defense attorneys, and misattributions of PTP as having been presented as trial evidence. Similar effects were obtained for negative and positive PTP. Delay significantly increased source-memory errors but did not influence guilt ratings. Defendant's credibility and juror ratings of prosecuting and defense attorneys significantly mediated the effect of PTP on guilt ratings. (PsycINFO Database Record (c) 2008 APA, all rights reserved).