Self‐referential complex trial protocol: Identifying criminals and witnesses in virtual reality scenarios

Abstract Purpose The item‐source Complex Trial Protocol (CTP) can effectively distinguish between guilty individuals and knowledgeable innocents. However, it is only suitable for theft cases and is prone to misjudging knowledgeable innocents as guilty. Therefore, it is necessary to develop other CTP paradigms that can distinguish the guilty from knowledgeable innocents. Methods This study designed a self‐referential CTP based on the difference in the degree of self‐referential processing of crime‐related information between the guilty and knowledgeable innocents. In a virtual reality scenario, the guilty participants completed a simulated murder, while the knowledgeable innocent participants witnessed the murder. In each trial of the self‐referential CTP, the first part randomly presented a probe (i.e. the victim's fac e) or an irrelevant stimulus (i.e. an unknown passer‐by's face ), and the second part randomly presented a target (i.e. the participant's own birthday ) or a non‐target (i.e. an irrelevant date ). All participants were required to make a response of ‘relevant to me’ or ‘irrelevant to me’ when they saw the faces and dates. The early posterior negativity (EPN), which indexes the degree of self‐referential processing, was measured. Results The results showed that the difference in EPN amplitude between irrelevant stimuli and probe after bootstrap resampling was significantly larger in the guilty group than in both the knowledgeable and unknowledgeable innocent groups. Conclusions The self‐referential CTP not only distinguished the guilty from knowledgeable innocents effectively (AUC = 0.90) but also discriminated the guilty from unknowledgeable innocents effectively (AUC = 0.88), correctly classifying knowledgeable innocents as innocent.