Are multiple-trial experiments appropriate for eyewitness identification studies? Accuracy, choosing, and confidence across trials

Abstract

Eyewitness identification experiments typically involve a single trial: a participant views an event and subsequently makes a lineup decision. Compared to this single-trial paradigm, multiple-trial designs are more efficient but significantly reduce ecological validity and may affect the strategies participants use to make lineup decisions. We examined the effects of a number of forensically-relevant variables (i.e., memory strength, type of disguise, degree of disguise, and lineup type) on eyewitness accuracy, choosing, and confidence across 12 target-present and 12 target-absent lineup trials (N = 349; 8,376 lineup decisions). Rates of correct rejections and choosing (across both target-present and -absent lineups) did not vary across the 24 trials as reflected by main effects or interactions with trial number. Trial number had a significant but trivial quadratic effect on correct identifications (OR = 0.99) and interacted significantly, but again trivially, with disguise type (OR = 1.00). Trial number did not significantly influence participants' confidence in correct identifications, confidence in correct rejections, or confidence in target-absent selections. Thus, multiple-trial designs appear to have minimal effects on eyewitness accuracy, choosing, and confidence. Researchers should consider using multiple-trial designs for conducting eyewitness identification experiments.