The Accuracy of Eyewitness Identification in Lineups

Cuny John Jay College Of Criminal Justice, New York NY

Investigators

Abstract

In criminal justice it is critical to identify guilty suspects accurately and avoid mistaken identification of innocent suspects. Consequently, best practices for constructing lineups provide that suspects should not stand out from other lineup members. Yet, data has established that criminal suspects with facial features matching an African phenotype are more likely to be chosen from a lineup than those with less Black or African features. This puts innocent suspects with African facial features at risk of misidentification, especially if lineups are constructed with suspects having more African facial features than other lineup members. This project examines whether certain facial features influence mistaken identification of innocent suspects and correct identifications of guilty ones. It thus will provide information designed to improve the accuracy of eyewitness identification, as well as guidance to judges, attorneys, and jurors when evaluating the identification of phenotypically African suspects. This project will involve a series of studies that examine how phenotypic facial features affect the accuracy of eyewitness identification. First, this project will test whether phenotypic bias (the tendency to view people with African features as criminal) in eyewitness identification is due to stereotype influence during the encoding of witnessed events, failures to use deliberative processes to correct for stereotype influence on judgments, or a combination of the two. This study also will test assumptions behind applied lineup theory, which suggests that automatic recognition processes lead to accurate identifications and deliberate processes introduce bias. These outcomes are in opposition to dual-process theories of stereotyping, which predict that phenotypic bias would influence judgments automatically and that deliberate processes would be required to correct for the biasing effects of stereotypes. Further studies in this project explore the interaction of phenotypic bias with factors that bias witnesses toward mistaken identifications. De-identified data from the project will be made publicly available. As a result, this project will test multiple theoretic foundations on lineups, as well as the potential mechanisms behind phenotypic bias, which may impair the accuracy of eyewitness identification.