Eyewitness Identification Can Be Studied in Social Contexts Online with Large Samples in Multi-Lab Collaborations

Abstract

Comments on an article by M. B. Kovera and A. J. Evelo (see record 2021-79725-001). Kovera and Evelo made a strong case for the value of studying eyewitness identification in social contexts. They argued that in recent years an influx of cognitive psychologists into the community of researchers studying eyewitness identification led to increased use of signal detection measures of identification performance. Kovera and Evelo particularly emphasized the use of receiver operating characteristic (ROC) curves, which demand large numbers of subjects. Improved understanding of statistical power may also have encouraged increases in sample size. Increased demand for large samples coincided with the rise of internet-based crowd-sourcing platforms that make it easy to test thousands of subjects in standardized procedures devoid of live person-to-person interaction. Kovera and Evelo called for caution in generalizing from such studies to real world situations in which motivated police officers work with witnesses. The authors agree with Kovera and Evelo that the field must do more to situate research in the social contexts so central to real-world eyewitness experiences.