The Effects of Post-Recall Feedback: Examining Witness Recall Quantity, Accuracy, and Confidence

Abstract

Most eyewitness identification protocols recommend withholding feedback after an identification has been made, at least until a measure of confidence can be gathered.Although much research has examined the impact of post-identification feedback on subsequent witness behavior and confidence, research addressing the importance of postrecall feedback remains largely incomplete. The current study examined the effects of post-recall feedback and question type on subsequent witness recall, confidence, and reports of view of the crime. In line with previous eyewitness identification research, it was predicted that participants receiving confirming post-recall feedback would be more confident in their prior recall compared to participants receiving neutral, no, or disconfirming feedback. One hundred and fifty-eight participants viewed a mock crime video of a robbery followed by an interview which included both open-ended and cued questions. Participants were then given either confirming, neutral, no, or disconfirming feedback and asked about their confidence and the clarity of their view of the perpetrator.Under the pretense that the recording equipment failed, participants were interviewed again using the same question format. The second interview was followed by a series of vii suggestive questions. After the second interview, participants were again asked about their confidence and self-reported quality of view of the perpetrator. Participant interviews were transcribed and scored for quantity, accuracy, and consistency by two blind, independent coders. Analyses revealed that feedback had a systematic impact on confidence such that participants who received confirming feedback were more confident in the overall accuracy of their prior memory accounts than those who received neutral, no, or disconfirming feedback and participants who received neutral or no feedback were more confident in the overall accuracy of their prior memory accounts compared to those in the disconfirming feedback condition. In line with previous eyewitness identification research, there was no significant relationship between recall accuracy and reported confidence. Results from the current study can be used to inform real-world investigative interviewers by highlighting the consequences of offering post-recall feedback.