ROC analysis of the verbal overshadowing effect: testing the effect of verbalisation on memory sensitivity
2016-01-28T14:35:16Z (GMT) by
This study investigated the role of memory sensitivity versus recognition criterion in the verbal overshadowing effect (VOE). Lineup recognition data were analysed using receiver operating characteristic analysis to separate the effects of verbalisation on memory sensitivity from criterion placement. Participants watched a short crime video, described the perpetrator’s facial features and then attempted a lineup identification. Description instructions were varied between participants. There was a standard (free report), forced (report everything) and warning (report accurate information) condition. Control participants did not describe the perpetrator. Memory sensitivity was greater in the control compared with the standard condition. Memory sensitivity was also greater in the warning condition than in the forced and standard conditions. Memory sensitivity did not differ across the forced and standard-description conditions, although a more conservative lineup decision standard was employed in the forced condition. These results, along with qualitative analyses of descriptions, support both retrieval-based and criterion- based explanations of the VOE.