Experts but not lay people adjust guilt beliefs after mock interviews with innocent suspects

Prior research indicated that guilt presumptive interviewers ask suspects more confirmatory than exploratory questions, and are less likely to change guilt beliefs post interview. This may harm investigative decision making. We build on these prior studies in two ways. First, we test whether prior guilt beliefs still affect post-interview decision making when interviewers use scripted non-accusatory open questions, and when suspects always gave a plausible alternative account for the evidence held. I.e. all suspects in our study were innocent. Second, most prior research has used student samples, so we compare the effects of pre-interview guilt beliefs in a sample of non-experts and experienced detectives. We found that pre-and post guilt beliefs correlated in the sample of non-experts (r = .59) but not experts (r = .08 ). We also found that guilt beliefs did not reduce pre- to post-interview for non-experts (p = .456), but did for experts (p = .027). These results tentatively show that experts can overcome prior assumptions if exculpatory evidence is provided by suspects in response to non-accusatory open questions. More clearly, we show the importance of using expert samples, where feasible.