Active exploration of faces in police lineups increases discrimination accuracy.

Abstract

Eyewitness identifications play a key role in the justice system, but eyewitnesses can make errors, often with profound consequences. We used findings from basic science and innovative technologies to develop and test whether a novel interactive lineup procedure, wherein witnesses can rotate and dynamically view the lineup faces from different angles, improves witness discrimination accuracy compared with a widely used procedure in laboratories and police forces around the world-the static frontal-pose photo lineup. No novel procedure has previously been shown to improve witness discrimination accuracy. In Experiment 1, participants (N = 220) identified culprits from sequentially presented interactive lineups or static frontal-pose photo lineups. In Experiment 2, participants (N = 8,507) identified culprits from interactive lineups that were either presented sequentially, simultaneously wherein the faces could be moved independently, or simultaneously wherein the faces moved jointly into the same angle. Sequential interactive lineups enhanced witness discrimination accuracy compared with static photo lineups, and simultaneous interactive lineups enhanced witness discrimination accuracy compared with sequential interactive lineups. These finding were true both when participants viewed suspects who were of the same or different ethnicity/ race as themselves. Our findings exemplify how basic science can be used to address the important applied policy issue on how best to conduct a police lineup and reduce eyewitness errors.

Child witness expressions of certainty are informative.

“…In other related work using interactive lineups, we have manipulated the encoding angle of a culprit and found that adults had better memory discrimination accuracy when they spent a longer proportion of time rotating the lineup faces to view the side of the culprit’s face that they had viewed at encoding (Colloff et al, 2020). Moreover, interactive lineups can substantially improve adult memory discrimination accuracy compared with traditional lineups which are composed of static photos of the lineup members facing the camera (Colloff et al, in press). Together, it seems that interacting for a longer length of time did not necessarily harm memory discrimination accuracy in our sample of children.…”

Three‐level meta‐analysis of the other‐race bias in facial identification

“…There were 83 articles, which qualified the inclusion criteria. From the 83 articles, 11 articles were removed because either they focused on cross‐race effects and did not provide any information of the other‐race bias effects for participants of each race (e.g., Colloff et al, 2021; Rothweiler et al, 2020), or because they did not provide necessary information to calculate effect sizes (e.g., Davis et al, 2020; Shafai & Oruc, 2018). The 87 articles selected from Lee's database (2019) also satisfied the above inclusion and exclusion criteria.…”

Legal system v. Eyewitness: The jury is still out on who is better able to reduce eyewitness error (variance).

“…Counterintuitively, two procedures that have comparable reliability can nevertheless differ dramatically in how well they correctly sort innocent and guilty suspects. For example, both simultaneous and sequential lineups seem to be equally reliable in that high-confidence suspect identifications result in similarly high probabilities of guilt, yet the simultaneous lineup typically has a higher ROC (e.g., see Colloff et al, in press; Mickes, 2015; Seale-Carlisle et al, 2019). The implication of this ROC advantage is that the simultaneous lineup would yield a larger number of reliable, high-confidence suspect identifications than the sequential lineup.…”