Eyewitness Identification: Victims versus Bystanders


Requests for reprints should be sent to Saul M. Kassin, Department of Psychology, Bronfman Science Center, Williams College, Williamstown, MA 01267.


The present study compared the eyewitness performance of crime victims and bystanders. Fifteen pairs of subjects participated in what they thought to be a risk-taking experiment that involved betting on the outcomes of a competitive game. In each session, the contest was interrupted when a confederate entered the room and stole the game money from one of the subjects while the other subject, seated across the table, watched. Whereas eight bystanders subsequently recognized the thief from a seven-picture photospread, not a single victim made an accurate identification (p < .001). On physical descriptions of the confederate, a witness sex × role interaction (p < .05) revealed that this victim-bystander effect held for female but not male subjects. Both arousal-based and attentional explanations of these data are considered. The findings are also discussed for their obvious practical implications for eyewitness testimony evidence.