Date of Degree


Document Type


Degree Name





Margaret Bull Kovera

Committee Members

Deryn Strange

Steven Penrod

Steven Young

Gary Wells

Subject Categories

Law | Social and Behavioral Sciences | Social Justice | Social Psychology


cross-race effect, eyewitness identification, lineups, policing, racial disparities, plea bargaining


There are clear racial disparities in the rates of wrongful convictions, with Black exonerees disproportionately represented among the population of those exonerated, in DNA and non-DNA exonerations alike (National Registry of Exonerations, 2022; Innocence Project, 2022). This racial disparity also exists for those exonerees who were wrongfully convicted, at least in part, because an eyewitness mistakenly identified them. For decades, when eyewitness scholars explored racial bias, they focused on the cross-race effect or own-race bias among eyewitnesses, a bias positing that witness performance suffers when a witness is asked to make an identification of a cross-race face (Lee & Penrod, 2022; Meissner & Brigham, 2001). However, given that most crime occurs within-race (NACJD, 2016), the cross-race effect cannot explain racial disparities in exoneration data. This work examined mechanisms that might contribute to racial disparities in exoneration data, with a particular focus on racial disparities in the base rate of culprit-present lineups. In Study 1, I explored whether racial disparities in evidence-based suspicion (i.e., evidence of guilt prior to placement in a lineup) provide a better explanation of racial disparities in exonerations based on eyewitness misidentification than the own-race bias in eyewitness identifications. To do so, I conducted a meta-analysis on 54 effect sizes extracted from 16 studies (1,503 individual participants) that tested whether there was an own-race bias in eyewitness identifications using a design that varied the race of both the witnesses and the target faces (Black vs. White). I also constructed two curves that plotted the prior probability of suspect guilt against the posterior probability of guilt: one if an identification were to be obtained for a Black suspect and one if an identification were to be obtained for a White suspect. Participants, irrespective of their race, were better able to discriminate among previously seen White than Black targets. However, the differential accuracy rates for identifications of White versus Black suspects were too small to explain racial disparities in exoneration data on their own, suggesting that racial disparities in evidence that police have against suspects before placing them in an identification procedure would likely explain more of the variance in racial disparities in mistaken identifications that lead to wrongful convictions.

In Study 2, I examined whether defense attorneys are less sensitive to prior evidence of guilt when the defendant is Black as opposed to White. To do so, I gave 316 defense attorneys NYPD-style case files that varied the strength of the pre-identification evidence (strong vs. weak), the race of the defendant (Black vs. White), and the race of the victim (Black vs. White). Attorneys made judgments that were sensitive to the base rate of guilt, but self-report measures demonstrated that they did not understand the extent to which the base rate of guilt influences the reliability of eyewitness evidence. Participants also rated the strength of the pre-identification evidence as stronger for Black than for White defendants. This research supports the conclusion that racially biased investigatory judgments may explain racial disparities in exoneration data based on eyewitness misidentifications. Additionally, although attorneys are intuitively sensitive to the strength of pre-identification evidence, they lack conscious awareness to this issue which may have implications for their ability to make race-neutral evaluations of pre-identification evidence.