Evaluating confessions in context: a policy-capturing study

Given the well-documented risks associated with confession evidence, we examined how confessions influence lay perceptions of culpability, alongside other common forms of evidence. We hypothesized that DNA and confession evidence would dominate judgments, with forensic expert, eyewitness, and motive evidence increasing guilt ratings but exerting comparatively less influence. We further anticipated that confessions would carry more weight when corroborated and be weakened by exculpatory evidence. Participants (N = 100) identified primarily as European/White (73%) and cisgender men (50%), aged 20–63 years (M = 39.97; SD = 11.78). Using a policy-capturing approach, participants evaluated 32 vignettes that varied the presence and nature (inculpatory vs. exculpatory) of five evidence types. Participants provided guilt ratings and rank-ordered each evidence type by importance. Results showed that DNA and confession evidence exerted the strongest influence on guilt ratings, with confession evidence emerging as a particularly potent predictor – even when contradicted by exculpatory DNA evidence. A modest association between self-reported importance and estimated evidence weighting suggests laypeople’s stated beliefs only partially align with how they integrate evidence. These findings reinforce concerns about the disproportionate weight assigned to confessions and clarify how evidence is weighted across evidentiary contexts – sometimes in ways that diverge from individuals’ stated beliefs.