The Impact of Race on Assessor Ability to Differentiate Accurate and Inaccurate Witness Identifications: Areas of Vulnerability, Bias, and Discriminatory Outcomes

Abstract

In this work we investigate potential weaknesses of lay assessors in differentiating true and false eyewitness identifications, through an examination of the influence of assessor, witness, and suspect race on assessor evaluations of eyewitness identifications made following observation of a mock crime. In an initial mock crime study (N= 209) we generate a set of same- and cross- race identifications, which are then assessed by participants in two experiments (N=1,144; N=675). Results support findings from existing mock juror work including showing that that assessors do not (or do not sufficiently) account for the race of a witness and target in assessing identifications, and that assessments can differ systematically based on assessor race. Importantly, results also show how these realities can impact overall patterns of accuracy in assessments of identifications. Specifically, results demonstrate a particular susceptibility to inaccuracy (resulting from a bias towards believing identifications are accurate) in evaluations of Black suspects by White witnesses, and superior ability of Black assessors (compared to White assessors) to differentiate accurate and inaccurate identifications of Black suspects. We also show that assessor weaknesses in accounting for race are not outweighed by contextual knowledge – a simple model accounting for witness and suspect race and confidence outperformed assessors in predicting identification accuracy. Findings highlight the need for particular scrutiny to be applied to assessments of identification decisions made by White witnesses in relation to Black lineups, and also underscore the importance of jury diversity to enhance the accuracy of assessments of evidence.