Evaluations of testimony of eyewitnesses who participated in multiple identification procedures

Laura Smalarz, Arizona State University
Amy Bradfield Douglass, Bates College
Angela Chang, Williams College

Eyewitnesses are sometimes asked to participate in multiple lineup procedures over the course of a criminal investigation, yet little is known about how evaluators perceive eyewitnesses who responded to an earlier identification procedure. In the current research, evaluators (N = 175) read a partial trial transcript in which the eyewitness testified to having participated in either 1 or 2 lineups over the course of the investigation. In the prior lineup conditions, the eyewitness made an identification or a rejection decision with either high or low confidence from the prior lineup, which did not contain the defendant, and later identified the defendant from a subsequent lineup. As hypothesized, making a prior identification of someone other than the defendant decreased perceptions of the eyewitness’s accuracy and led to fewer guilty verdicts. Contrary to our hypothesis, however, confidently rejecting a prior lineup that did not contain the defendant did not increase perceptions of the eyewitness’s accuracy. In Experiment 2 (N = 233), we tested whether presenting eyewitness identification evidence via videotape—rather than solely through the eyewitness’s testimony as in Experiment 1—strengthens the effect of a prior rejection decision on evaluations of eyewitness accuracy. As hypothesized, evidence of a prior rejection increased perceptions of an eyewitness’s accuracy, but only when the identification evidence was presented via videotape. These findings illustrate that evaluators incorporate information about witnesses’ prior lineup decisions into their legal judgments and that evidence of a prior identification constitutes Brady material that should be disclosed to the defense.

eyewitness identification, eyewitness testimony, evaluations of eyewitnesses, eyewitness confidence, multiple lineups

“In many cases of wrongful conviction resulting from mistaken eyewitness identification testimony, the eyewitnesses made their identifications after having already participated in an earlier identification procedure. […] An analysis of the first 161 DNA exoneration cases involving eyewitness misidentification indicated that in 40% of the cases, the eyewitness initially identified a filler, another suspect, or rejected an initial identification procedure before ultimately identifying the defendant in a subsequent procedure.” (p. 10)

“Despite the prevalence of cases in which eyewitness identification evidence is obtained following a previous identification attempt, little is known about how people evaluate eyewitnesses who have participated in multiple identification procedures.” (p. 10)

“An eyewitness’s response to a police lineup can be classified into one of three categories: a suspect identification, a filler identification, or a lineup rejection. When the eyewitness identifies the suspect, the identification constitutes incriminating evidence against the suspect and the suspect can be prosecuted on the basis of that evidence. In such cases, police are unlikely to conduct a subsequent identification procedure with a new suspect. When the eyewitness identifies a filler or rejects the lineup, however, it suggests that the suspect may be innocent, and police must continue investigating the case. If the police identify a new possible suspect, they might conduct a subsequent identification procedure with the eyewitness. Field data from real criminal investigations indicate that the use of multiple identification procedures may be quite common: In as many as 74% of police lineups, the eyewitness either identifies a filler or rejects the lineup rather than identifying the suspect.” (p. 11)

“The focus of the current research, therefore, is how evaluators perceive identification testimony from an eyewitness who made an identification or rejection decision from a previous lineup before identifying the defendant from a subsequent lineup. Consistent with scenarios in which the police pursue a new suspect following the nonidentification of an initial suspect, the initial lineup in the current research did not contain the eventually-identified defendant.” (p. 11)

“Prevailing legal intuition aligns with the idea that witnesses who make an inaccurate initial identification are not to be trusted in subsequent identification procedures whereas witnesses who reject an initial lineup can still be considered reliable. […] The most common type of erroneous initial identification occurs when an eyewitness identifies a known-innocent lineup filler, which occurs in nearly one in four cases involving real eyewitnesses. Less common are situations in which the witness initially identifies the police suspect and that suspect is later determined to be innocent of the crime.” (p. 11)

“Nevertheless, in both scenarios, the eyewitness has made an inaccurate identification and thus is unlikely to provide reliable evidence from a subsequent lineup. By contrast, witnesses who (correctly) reject an initial lineup have been shown to perform similarly on a subsequent identification task to witnesses who were not exposed to an initial lineup.” (p. 11)

“Despite the empirically supported assumption that witnesses who make an inaccurate identification from an initial lineup are unreliable, there are numerous examples of cases in which eyewitnesses who made a previous inaccurate identification were nonetheless permitted to testify in court about a subsequent identification.” (p. 11)

“Understanding how prior lineup decisions affect evaluators’ legal judgments has important implications for legal policy for at least two reasons. First, if a prior identification harms an eyewitness’s credibility in the eyes of evaluators, then evidence of a prior identification decision would be considered Brady material— evidence in the possession of the state or government that could favor the defense. As a result, withholding evidence of a witness’s prior identification from the defense would be in violation of constitutional law.” (p. 12)

“Second, if a high-confidence prior rejection decision enhances perceptions of an eyewitness’s credibility, then prosecutors might be eager to introduce evidence of eyewitnesses’ prior rejections at trial. Of course, information about witnesses’ prior rejection decisions will generally be available only if eyewitness identification decisions are faithfully documented by police. Prosecutors have notoriously been resistant to eyewitness identification reform, but might become amenable to reform efforts that produce evidence that could be used to their benefit in court.” (p. 12)

“We tested three hypotheses in our first experiment. First, we hypothesized that an eyewitness who identified someone from a previous lineup that was known to not contain the defendant is perceived as less accurate than an eyewitness who did not participate in a prior lineup procedure (H1). Second, we predicted that expressing high (vs. low) confidence in a prior lineup decision has different effects on perceptions of subsequent identification accuracy depending on the witness’s initial decision, with high confidence decreasing perceptions of accuracy for an eyewitness who made a prior identification and increasing perceptions of accuracy for an eyewitness who rejected a prior lineup (H2). Finally, we predicted that a high-confidence prior rejection decision boosts perceptions of eyewitness accuracy relative to that of an eyewitness who was not exposed to a prior lineup (H3).” (p. 12)

“Participants were randomly assigned to a 2 (prior lineup decision: identification vs. rejection) X 2 (prior decision confidence: high vs. low) between-subjects design with a hanging no-prior lineup control group in which the eyewitness was exposed to only one lineup. Our primary dependent measures were participants’ perceptions of the eyewitness’s accuracy and their verdict decisions.” (pp. 12–13)

“Participants were 175 Amazon Mechanical Turk workers […] Participants’ age ranges were reported as follows: 18 29 years (n = 74), 30–39 years (n = 51), 40–49 years (n = 38), 50–59 years (n = 8), and 60 years or older (n = 4). Sixteen participants self-identified as Asian or Pacific Islander, 13 as Black or African American, 10 as Hispanic or Latino, one as Native American, 127 as White, and eight as multiracial or other. Participants included 107 participants who identified as male, 67 who identified as female, and one who identified as other. The majority of participants had never served on a jury before (n = 154) nor had been an eyewitness to a crime (n = 138).” (p. 13)

“Consistent with the literature on perceptions of confidence calibration, an eyewitness’s prior identification tended to be more damaging to the eyewitness’s credibility when it was made with high confidence than when it was made with low confidence, whereas greater confidence tended to increase perceptions of accuracy when the eyewitness rejected the previous lineup.” (p. 15)

“Contrary to our third hypothesis, however, an eyewitness who made a highly confident prior rejection decision was not perceived as more accurate compared to an eyewitness who was not exposed to a prior lineup. A possible explanation for this finding is that evaluators simply do not know how to interpret an eyewitness’s prior rejection decision because there is no clear indication as to whether or not the prior rejection was the correct response. As a result, evaluators might simply disregard the evidence of the prior rejection decision when evaluating the likely accuracy of the eyewitness’s subsequent identification of the defendant. Another possibility, however, is that an eyewitness’s prior rejection decision influences evaluators’ perceptions of the eyewitness’s accuracy only under certain conditions.” (p. 15)

“Our aim in Experiment 2 was twofold: First, we tested the hypothesis that presenting a witness’s prior rejection decision via videotape boosts evaluators’ perceptions of eyewitness accuracy to a greater extent than when the information about the witness’s prior rejection is presented solely via the eyewitness’s testimony. Second, we sought to replicate our other findings from Experiment 1 under conditions in which the eyewitness evidence is presented via videotape rather than solely through the eyewitness’s testimony.” (p. 16)

“Participants were randomly assigned to a 3 (prior lineup: no prior lineup, prior identification, prior rejection) X 2 (identification evidence: testimony only vs. testimony X videotape) between-subjects design. In the testimony X videotape conditions, all identification procedures were described in the transcript and shown on video; in the testimony-only conditions, the identification procedures were described through the witness’s testimony only. To maximize statistical power, the eyewitness was always 100% confident in her decision from the prior lineup, consistent with the high-confidence conditions in Experiment 1.” (p. 16)

“Participants were 233 Amazon Mechanical Turk workers […] Participants’ age ranges were reported as follows: 18–29 years (n = 94), 30–39 years (n = 87), 40–49 years (n = 30), 50–59 years (n = 16), and 60 years or older (n = 6). Seventeen participants self-identified as Asian or Pacific Islander, 20 as Black or African American, 12 as Hispanic or Latino, one as Native American, 179 as White, and four as multiracial or other. Participants included 137 participants who identified as male, 95 who identified as female, and one who identified as other. The majority of participants had never served on a jury before (n = 195) nor had been an eyewitness to a crime (n = 183).” (p. 16)

“As in Experiment 1, participants perceived the eyewitness as less accurate and were less likely to render a guilty verdict upon learning that she had identified someone other than the defendant from a previous lineup. Importantly, these patterns of results held in the testimony _ videotape conditions in Experiment 2 as well as in the testimony-only conditions, thereby replicating and extending the findings from Experiment 1. Furthermore, Experiment 2 provided evidence that there are conditions under which a high-confidence prior rejection decision can boost perceptions of an eyewitness’s accuracy—namely, when the identification evidence is presented via videotape.” (p. 18)

“Although the eyewitness’s prior rejection increased evaluators’ perceptions of eyewitness accuracy when presented via videotape, it did not correspondingly influence guilty verdicts. This finding suggests that even when a witness’s prior rejection decision is presented via videotape at trial, its effects on evaluators’ judgments may still be limited. One possible explanation for these limited effects of the prior rejection could be that presenting a videotape of the witness’s prior rejection has other, independent effects on evaluators’ judgments.” (p. 18)

“The current research demonstrates that a witness’s prior identification of someone other than the defendant harms an eyewitness’s credibility. This finding is particularly consequential because it suggests that evidence of a witness’s prior identification decision should be considered “Brady material” and that withholding evidence of a witness’s prior identification from the defense would thus be in violation of constitutional law. Such a finding may prove useful to defense attorneys, who could leverage these data to strengthen their Brady motions soliciting evidence of previously conducted identification procedures.” (p. 20)

“Our results also indicated that evidence of a witness’s prior rejection has a relatively weak effect on evaluators’ judgments unless it is presented via videotape, in which case a prior rejection can boost perceptions of an eyewitness’s accuracy. This latter finding may be of particular interest to prosecuting attorneys, who have a reputation for resisting eyewitness identification reforms. If prosecuting attorneys were aware that presenting videotaped evidence of a witness’s prior lineup rejection can help enhance a witness’s credibility at trial, they might be more amenable to the instatement of a videotaping policy for all eyewitness identification procedures.” (p. 20)

“Unfortunately, however, the available empirical data on subsequent lineup performance suggest that witnesses who reject an initial lineup perform no better on a subsequent lineup than do witnesses who did not participate in an initial lineup. Therefore, the benefit that our findings might bring to the endorsement of video-recording practices could come at a cost of overinflating jurors’ belief of eyewitnesses who rejected a previous lineup.” (p. 20)

“One limitation of the current research is that evidence of an eyewitness’s prior lineup rejection was always presented by the prosecuting attorney, whereas evidence of an eyewitness’s prior identification was always presented by the defense attorney. We chose to introduce the prior lineup evidence in this way because we believe that it captures what is most likely to occur in real cases, in which the prosecution typically introduces incriminating evidence and the defense typically introduces exculpatory evidence.” (p. 19)

“Another limitation of the current research is that our materials consisted of written trial transcripts in which the eyewitness evidence featured prominently because the transcripts were relatively brief and centered on the testimony of the eyewitness. Given that scant research has tested whether presenting evidence about an eyewitness’s prior lineup decision can influence evaluations of a subsequent identification, we chose to use materials that focused specifically on that aspect of a trial.” (p. 19)

“Concerns about our use of a trial transcript rather than a more realistic trial simulation are alleviated somewhat by research suggesting that whether trials are presented in a written transcript or in a more realistic format does not have consistent effects on evaluators’ judgments. Even so, the use of more varied materials in future research on perceptions of prior lineup decisions will strengthen the generalizability of this nascent line of research.” (p. 19)

As always, please join the discussion below if you have thoughts or comments to add!