Closing the Eyes Increases Accuracy but Decreases Confidence in Eyewitness Recall

Annelies Vredeveldt, University of York; James D. Sauer, University of Portsmouth

Recent research has shown that closing the eyes during an investigative interview can increase the amount and accuracy of eyewitness testimony. The present two experiments used a calibration approach to assess the effect of eye-closure on eyewitness confidence. Participants watched violent and non-violent video clips and were asked 20 specific questions about each video clip. They provided confidence ratings for each answer, on a scale of 1 to 5 in Experiment 1 and on a scale of 0 to 100% in Experiment 2. Experiment 1 showed that the instruction to close the eyes during the interview (compared to no instruction) reduced mean confidence irrespective of accuracy, and impaired confidence-accuracy (CA) calibration as a result of underconfidence. In Experiment 2, two interview conditions low in distraction (looking at a blank screen or closing the eyes) were compared to two interview conditions high in distraction (visual or auditory distractions). Interview condition did not affect mean confidence, but participants in the low-distraction conditions again displayed underconfidence in terms of CA calibration. The findings suggest that witnesses who are not exposed to sensory distractions during the interview fail to adjust their confidence ratings upwards in line with their superior accuracy (compared to witnesses exposed to distractions), resulting in underconfidence. The findings might be explained by an increased awareness of the difficulty of the retrieval task. From an applied perspective, the underconfidence associated with eye-closure is arguably preferable to the overconfidence associated with certain other interviewing techniques, such as hypnosis.