Meetings Make Evidence? An Experimental Study of Collaborative and Individual Recall of a Simulated Police Interrogation

An audio recording of a simulated police interrogation of a woman alleging rape was played to 58 subjects, who subsequently recalled and answered questions about it in one of three experimental conditions: individual, dyadic, and four-person group. Dyads and groups were required to agree on all responses. A macropropositional analysis of the interrogation was used to classify propositions in each recall protocol in terms of their correspondence to those of the original. Although groups produced twice as many "correct" propositions as did individuals, the total number of recall propositions, and reconstructive and confusional errors, did not vary between conditions, and individuals gave nearly five times as many metastatements as did groups. This analysis and further analyses of propositions correctly recalled suggest that groups gave conventional accounts of the evidence in contrast with individuals' evaluations of the interrogation. Groups and dyads were characteristically overconfident about wrong answers.