The effect of cognitive schema on eyewitness memory

Abstract

This study investigated the effect of cognitive schema, gender and type of information recalled on participants' memory of a video-recorded staged hijacking event. The results showed that, overall, participants who were exposed to the 'schema-inconsistent' video recalled significantly more accurate information than participants who were exposed to schema-consistent information (F (1, 58) = 14.671, P < 0.001). Gender did not significantly affect overall memory performance (F (1, 58) = 0.460, p > 0.05) nor did gender alone interact significantly with any other variable investigated in the study. However, there were significant main effects of the four types of information recalled; namely, hijacker details, victim details, time estimations, scene description and event sequencing. First, with respect to recall of hijacker details, participants were most accurate in their recall of the physical description of the hijacker(s), followed by the verbalisation(s) of the hijacker(s). Least accurate were participants' recall of the behaviour of the hijacker(s) (F (2, 58) = 16.394, p < 0.001). Second, with respect to victim details participants were most accurate in recalling the physical description of the victim(s) followed by the behaviour of the victim(s) and then lastly, the verbalisations of the victim(s) (F (2, 58) = 73.271, p < 0.001). Third, participants were significantly more accurate in their estimations of the time of day than they were in their estimation of event duration, which was consistently overestimated (F (1, 58) = 116.769, p < 0.001). In terms scene description and event sequencing, participants were significantly better at describing the scene than describing the sequence of events (F (1, 58) = 7.21, p < 0.01). With regards to interactions, condition interacted significantly with the recall of hijacker details (F (2, 58) = 5.713, p < 0.005). Recall of the schema-inconsistent group was better than that of the schema-consistent group on verbalisation(s) of the hijacker(s) and physical description of the hijacker(s) but the same for the behaviour of the hijacker(s). Also, a significant condition X description of scene X event sequencing interaction was found (F (1, 58) = 10.771, p < 0.0021). The schema-inconsistent group performed better than the schema-consistent group in terms of describing the scene but the two groups did not differ significantly in their recall of the sequence the event. The theoretical and practical implications of the findings from the study are presented and discussed. Also, recommendations for future research are made.