Regularities in eyewitness identification

The cognitive science of eyewitness memory

Recent insights from cognitive science have reshaped our understanding of the reliability of eyewitness memory. Many believe that eyewitness memory is unreliable, but a better way of thinking is that eyewitness memory, like other types of forensic evidence, can be contaminated. Because contaminated evidence yields unreliable results, the focus should be placed on testing uncontaminated memory evidence collected early in a police investigation. The recent application of theories, principles, and methods from cognitive science has revealed that, both in the laboratory and in the real world, the first test of uncontaminated memory provides much more reliable information than was previously thought. Moreover, and crucially, this reliable but often-ignored evidence frequently points in the direction of a convicted defendant's innocence.

Identification performance across the life span: Lineups and the reaction time-based Concealed Information Test

Cognitive and social factors can deteriorate eyewitness identification performance in children and older adults. An identification procedure that mitigates the effect of such factors could be beneficial for child and older adult witnesses. In a field experiment, we mapped identification performance in a large community sample (N = 1239) across the lifespan (ages 6-79 years) for two different identification procedures: classic lineups and reaction time-based Concealed Information Test (RT-CIT). Visitors of a science museum or science fair witnessed a recorded mock theft and then took either a classic lineup, or the RT-CIT. Young adults (18-35-year-olds) outperformed younger and older age groups in lineup performance. The RT-CIT showed a moderate capacity to diagnose face recognition and absence of recognition in the target-absent condition. Age did not affect identification with the RT-CIT. However, children were often not able to follow the RT-CIT instructions, leading to a large number of exclusions. A direct comparison of lineup vs. RT-CIT performance showed that children and adolescents showed better identification performance in RT-CIT than lineups. For young adults, there was no difference between the two procedures. The trend turned around at mid-adult age who showed better identification performance when they were given a lineup compared to an RT-CIT. These findings suggest that the RT-CIT might be considered an alternative identification procedure for children and adolescents, offering protection for innocent suspects.

New Insights on Expert Opinion About Eyewitness Memory Research

Experimental psychologists investigating eyewitness memory have periodically gathered their thoughts on a variety of eyewitness memory phenomena. Courts and other stakeholders of eyewitness research rely on the expert opinions reflected in these surveys to make informed decisions. However, the last survey of this sort was published more than 20 years ago, and the science of eyewitness memory has developed since that time. Stakeholders need a current database of expert opinions to make informed decisions. In this article, we provide that update. We surveyed 76 scientists for their opinions on eyewitness memory phenomena. We compared these current expert opinions to expert opinions from the past several decades. We found that experts today share many of the same opinions as experts in the past and have more nuanced thoughts about two issues. Experts in the past endorsed the idea that confidence is weakly related to accuracy, but experts today acknowledge the potential diagnostic value of initial confidence collected from a properly administered lineup. In addition, experts in the past may have favored sequential over simultaneous lineup presentation, but experts today are divided on this issue. We believe this new survey will prove useful to the court and to other stakeholders of eyewitness research.

Diagnosing eyewitness identifications with reaction time‑based Concealed Information Test: the effect of viewpoint congruency between test and encoding

Mistaken eyewitness identifications continue to be a major contributor to miscarriages of justice. Previous experiments have suggested that implicit identification procedures such as the Concealed Information Test (CIT) might be a promising alternative to classic lineups when encoding conditions during the crime were favorable. We tested this idea by manipulating view congruency (frontal vs. profile view) between encoding and test. Participants witnessed a videotaped mock theft that showed the thief and victim almost exclusively from frontal or profile view. At test, viewing angle was either congruent or incongruent with the view during encoding. We tested eyewitness identification with the RT-CIT (N = 74), and with a traditional simultaneous photo lineup (N = 97). The CIT showed strong capacity to diagnose face recognition (d = 0.91 [0.64; 1.18]), but unexpectedly, view congruency did not moderate this effect. View congruency moderated lineup performance for one of the two lineups. Following these unexpected findings, we conducted a replication with a stronger congruency manipulation and larger sample size. CIT (N = 156) showed moderate capacity to diagnose face recognition (d = 0.63 [0.46; 0.80]) and now view congruency did moderate the CIT effect. For lineups (N = 156), view congruency again moderated performance for one of the two lineups. Capacity for diagnosing face recognition was similar for lineups and RT-CIT in our first comparison, but much stronger for lineups in our second comparison. Future experiments might investigate more conditions that affect performance in lineups vs. the RT-CIT differentially.

The effects of lineup size on the processes underlying eyewitness decisions

Here we apply the two-high threshold eyewitness identification model to identify the effects of lineup size on the detection-based and non-detection-based processes underlying eyewitness decisions. In Experiment 1, lineup size was manipulated by showing participants simultaneous or sequential lineups that contained either three or six persons. In Experiment 2, the lineups contained either two or five persons. In both experiments, the culprit was better detected in smaller than in larger lineups. Furthermore, participants made fewer guessing-based selections in smaller than in larger lineups. However, guessing-based selection in larger lineups was not increased to a level sufficient to offset the effect of increased protection of suspects in larger lineups due to the fact that the guessing-based selections that occur are distributed across more persons. The results show that increasing the lineup size causes several changes in the detection-based and non-detection-based processes underlying eyewitness decisions.

Diagnosing eyewitness identifications with reaction time-based concealed information test: the effect of observation time

Eyewitness identification procedures rely heavily on explicit identification from lineups. Lineups have been criticized because they have a considerable error rate. We tested the potential of implicit identifications in a Concealed Information Test (CIT) as an alternative. Previous experiments have suggested that implicit identification procedures might be suited when viewing conditions were favorable. In two experiments, mock eyewitnesses (Ns = 94, 509) witnessed a videotaped mock theft with longer or shorter observation time. We derived eyewitness identifications either implicitly from CIT reaction times or explicitly from simultaneous photo lineups. In Experiment 2, we also manipulated perpetrator presence. In both experiments, the perpetrator-present CIT showed capacity to diagnose face recognition, with large effect sizes (d_E1 = 0.85 [0.51; 1.18]; d_E2 = 0.74 [0.52; 0.96]), as expected. Unexpectedly, no moderation by observation time was found. In line with our hypothesis, no CIT effect emerged in the perpetrator-absent condition, indicating the absence of recognition (d_E2 = 0.02 [- 0.17; 0.20]). We found no compelling evidence that one method would outperform the other. This work adds to accumulating evidence that suggests that, under favorable viewing conditions and replication provided, the RT-CIT might be diagnostic of facial recognition, for example when witnesses are hesitant of making an explicit identification. Future work might investigate conditions that affect performance in one, but not the other identification method.

Evaluating earwitness identification procedures: adapting pre-parade instructions and parade procedure

Voice identification parades can be unreliable, as earwitness responses are error-prone. In this paper we tested performance across serial and sequential procedures, and varied pre-parade instructions, with the aim of reducing errors. The participants heard a target voice and later attempted to identify it from a parade. In Experiment 1 they were either warned that the target may or may not be present (standard warning) or encouraged to consider responding "not present" because of the associated risk of a wrongful conviction (strong warning). Strong warnings prompted a conservative criterion shift, with participants less likely to make a positive identification regardless of whether the target was present. In contrast to previous findings, we found no statistically reliable difference in accuracy between serial and sequential parades. Experiment 2 ruled out a potential confound in Experiment 1. Taken together, our results suggest that adapting pre-parade instructions provides a simple way of reducing the risk of false identifications.

Efeitos do Alinhamento Justo e Similaridade de Rostos no Reconhecimento de Pessoas

Resumo Um falso reconhecimento de uma pessoa pode levar à condenação de um inocente. Um método efetivo de diminuir o falso reconhecimento é por meio do alinhamento, procedimento no qual o suspeito é apresentado em conjunto com outras pessoas - fillers (não suspeitos similares ao suspeito). Em um experimento foi comparado o desempenho de testemunhas em alinhamentos nos quais fillers apresentavam moderada ou alta similaridade em relação ao suspeito. Independentemente do grau de similaridade, suspeitos foram identificados com maior frequência que suspeitos inocentes e do que fillers, e fillers foram reconhecidos em maior frequência do que suspeitos inocentes. A similaridade entre fillers e suspeito não teve efeito na probabilidade de reconhecimento do suspeito, seja ele culpado ou inocente. Os resultados são discutidos à luz de teorias acerca do efeito de similaridade de fillers e implicações dos resultados para o sistema de justiça brasileiro.

The distinction between discriminability and reliability in forensic science

Forensic science plays an increasingly important role in the criminal justice system; yet, many forensic procedures have not been subject to the empirical scrutiny that is expected in other scientific disciplines. Over the past two decades, the scientific community has done well to bridge the gap, but have likely only scratched the tip of the iceberg. We offer the discriminability-reliability distinction as a critical framework to guide future research on diagnostic-testing procedures in the forensic science domain. We argue that the primary concern of the scientist ought to be maximizing discriminability and that the primary concern of the criminal justice system ought to be assessing the reliability of evidence. We argue that Receiver Operating Characteristic (ROC) analysis is uniquely equipped for determining which of two procedures or conditions has better discriminability and we also demonstrate how estimates of reliability can be extracted from this Signal Detection framework.

Testing a potential alternative to traditional identification procedures: Reaction time-based concealed information test does not work for lineups with cooperative witnesses

Direct eyewitness identification is widely used, but prone to error. We tested the validity of indirect eyewitness identification decisions using the reaction time-based concealed information test (CIT) for assessing cooperative eyewitnesses' face memory as an alternative to traditional lineup procedures. In a series of five experiments, a total of 401 mock eyewitnesses watched one of 11 different stimulus events that depicted a breach of law. Eyewitness identifications in the CIT were derived from longer reaction times as compared to well-matched foil faces not encountered before. Across the five experiments, the weighted mean effect size d was 0.14 (95% CI 0.08-0.19). The reaction time-based CIT seems unsuited for testing cooperative eyewitnesses' memory for faces. The careful matching of the faces required for a fair lineup or the lack of intent to deceive may have hampered the diagnosticity of the reaction time-based CIT.

Reaction time-based Concealed Information Test in eyewitness identification is moderated by picture similarity but not eyewitness cooperation

The reaction time-based Concealed Information Test (RT-CIT) has high validity in assessing recognition of critical information. Findings on its usefulness for diagnosing face recognition in eyewitnesses are inconsistent. Experiment 1 (N = 82) tested whether closely matching the faces of the probes and irrelevants, as required for a fair lineup, undermines RT-CIT usefulness. Preregistered Experiments 2a and 2b (Ns = 48), tested the role of eyewitness cooperativeness for RT-CIT validity. All participants watched a mock crime video and then completed an RT-CIT. As expected, the usefulness of the RT-CIT was moderated by picture similarity, with better detection for non-matched faces. Unexpectedly, eyewitness cooperation (conceal vs. reveal recognition), did not affect the validity of the RT-CIT. A large CIT effect observed in Experiment 2b further suggested that-even with matched faces-the RT-CIT might be of use when encoding conditions during the crime were favorable. Cases in which witnesses are unwilling or afraid to make an explicit identification might concern another possible application.

Matching novel face and voice identity using static and dynamic facial images

Research investigating whether faces and voices share common source identity information has offered contradictory results. Accurate face-voice matching is consistently above chance when the facial stimuli are dynamic, but not when the facial stimuli are static. We tested whether procedural differences might help to account for the previous inconsistencies. In Experiment 1, participants completed a sequential two-alternative forced choice matching task. They either heard a voice and then saw two faces or saw a face and then heard two voices. Face-voice matching was above chance when the facial stimuli were dynamic and articulating, but not when they were static. In Experiment 2, we tested whether matching was more accurate when faces and voices were presented simultaneously. The participants saw two face-voice combinations, presented one after the other. They had to decide which combination was the same identity. As in Experiment 1, only dynamic face-voice matching was above chance. In Experiment 3, participants heard a voice and then saw two static faces presented simultaneously. With this procedure, static face-voice matching was above chance. The overall results, analyzed using multilevel modeling, showed that voices and dynamic articulating faces, as well as voices and static faces, share concordant source identity information. It seems, therefore, that above-chance static face-voice matching is sensitive to the experimental procedure employed. In addition, the inconsistencies in previous research might depend on the specific stimulus sets used; our multilevel modeling analyses show that some people look and sound more similar than others.

Costs and Benefits of Eyewitness Identification Reform: Psychological Science and Public Policy

Psychological science has come to play an increasingly important role in the legal system by informing the court through expert testimony and by shaping public policy. In recent years, psychological research has driven a movement to reform the procedures that police use to obtain eyewitness identification evidence. This reform movement has been based in part on an argument suggesting that recommended procedures reduce the risk of false identifications with little or no reduction in the rate of correct identifications. A review of the empirical literature, however, challenges this no-cost view. With only one exception, changes in eyewitness identification procedures that reduce the risk of false identification of the innocent also reduce the likelihood of correct identification of the guilty. The implication that criminals may escape prosecution as a result of procedures implemented to protect the innocent makes policy decisions far more complicated than they would otherwise be under the no-cost view. These costs (correct identifications lost) and benefits (false identifications avoided) are discussed in terms of probative value and expected utility.

Configural and component processing in simultaneous and sequential lineup procedures

Configural processing supports accurate face recognition, yet it has never been examined within the context of criminal identification lineups. We tested, using the inversion paradigm, the role of configural processing in lineups. Recent research has found that face discrimination accuracy in lineups is better in a simultaneous compared to a sequential lineup procedure. Therefore, we compared configural processing in simultaneous and sequential lineups to examine whether there are differences. We had participants view a crime video, and then they attempted to identify the perpetrator from a simultaneous or sequential lineup. The test faces were presented either upright or inverted, as previous research has shown that inverting test faces disrupts configural processing. The size of the inversion effect for faces was the same across lineup procedures, indicating that configural processing underlies face recognition in both procedures. Discrimination accuracy was comparable across lineup procedures in both the upright and inversion condition. Theoretical implications of the results are discussed.

Nosewitness identification: effects of negative emotion

Every individual has a unique body odor (BO), similar to a fingerprint. In forensic research, identification of culprit BOs has been performed by trained dogs, but not by humans. We introduce the concept of nosewitness identification and present the first experimental results on BO memory in witness situations involving violent crimes. Two experiments indicated that BO associated with male characters in authentic videos could later be identified in BO lineup tests well above chance. Moreover, culprit BO in emotional crime videos could be identified considerably better than the BO of a male person in neutral videos. This indicates that nosewitness identification benefits from emotional encoding. Altogether, the study testifies to the virtue of body odor as a cue to identify individuals observed under negative emotion.

Probative value of absolute and relative judgments in eyewitness identification

It is well-accepted that eyewitness identification decisions based on relative judgments are less accurate than identification decisions based on absolute judgments. However, the theoretical foundation for this view has not been established. In this study relative and absolute judgments were compared through simulations of the WITNESS model (Clark, Appl Cogn Psychol 17:629-654, 2003) to address the question: Do suspect identifications based on absolute judgments have higher probative value than suspect identifications based on relative judgments? Simulations of the WITNESS model showed a consistent advantage for absolute judgments over relative judgments for suspect-matched lineups. However, simulations of same-foils lineups showed a complex interaction based on the accuracy of memory and the similarity relationships among lineup members.

Sequential lineup laps and eyewitness accuracy

Police practice of double-blind sequential lineups prompts a question about the efficacy of repeated viewings (laps) of the sequential lineup. Two laboratory experiments confirmed the presence of a sequential lap effect: an increase in witness lineup picks from first to second lap, when the culprit was a stranger. The second lap produced more errors than correct identifications. In Experiment 2, lineup diagnosticity was significantly higher for sequential lineup procedures that employed a single versus double laps. Witnesses who elected to view a second lap made significantly more errors than witnesses who chose to stop after one lap or those who were required to view two laps. Witnesses with prior exposure to the culprit did not exhibit a sequential lap effect.

Exploring the sequential lineup advantage using WITNESS

Advocates claim that the sequential lineup is an improvement over simultaneous lineup procedures, but no formal (quantitatively specified) explanation exists for why it is better. The computational model WITNESS (Clark, Appl Cogn Psychol 17:629-654, 2003) was used to develop theoretical explanations for the sequential lineup advantage. In its current form, WITNESS produced a sequential advantage only by pairing conservative sequential choosing with liberal simultaneous choosing. However, this combination failed to approximate four extant experiments that exhibited large sequential advantages. Two of these experiments became the focus of our efforts because the data were uncontaminated by likely suspect position effects. Decision-based and memory-based modifications to WITNESS approximated the data and produced a sequential advantage. The next step is to evaluate the proposed explanations and modify public policy recommendations accordingly.

The effect of retention interval on the confidence-accuracy relationship for eyewitness identification

Recent research using a calibration approach indicates that eyewitness confidence assessments obtained immediately after a positive identification decision provide a useful guide as to the likely accuracy of the identification. This study extended research on the boundary conditions of the confidence-accuracy (CA) relationship by varying the retention interval between encoding and identification test. Participants (N = 1,063) viewed one of five different targets in a community setting and attempted an identification from an 8-person target-present or -absent lineup either immediately or several weeks later. Compared to the immediate condition, the delay condition produced greater overconfidence and lower diagnosticity. However, for choosers at both retention intervals there was a meaningful CA relationship and diagnosticity was much stronger at high than low confidence levels.

Repeated eyewitness identification procedures: memory, decision making, and probative value

Two experiments examined the effects of multiple identification procedures on identification responses, confidence, and similarity relationships. When the interval between first and second identification procedures was long (Experiment 1), correct and false identifications increased, but the probative value of a suspect identification changed little; consistent witnesses were more confident than inconsistent witnesses; and the similarity relationships between suspect and foils were unchanged. When the interval between first and second identification procedures was short (Experiment 2), suspect identification rates changed little, but foil identifications increased significantly; confidence for all identifications increased; consistent witnesses were more confident than inconsistent witnesses; and similarity relationships changed such that witnesses were less likely to identify the suspect as being the best match to the perpetrator.

Eyewitness identification evidence and innocence risk

It is well known that the frailties of human memory and vulnerability to suggestion lead to eyewitness identification errors. However, variations in different aspects of the eyewitnessing conditions produce different kinds of errors that are related to wrongful convictions in very different ways. We present a review of the eyewitness identification literature, organized around underlying cognitive mechanisms, memory, similarity, and decision processes, assessing the effects on both correct and mistaken identification. In addition, we calculate a conditional probability we call innocence risk, which is the probability that the suspect is innocent, given that the suspect was identified. Assessment of innocence risk is critical to the theoretical development of eyewitness identification research, as well as to legal decision making and policy evaluation. Our review shows a complex relationship between misidentification and innocence risk, sheds light on some areas of controversy, and suggests that some issues thought to be resolved are in need of additional research.