Jurors' Use of Naked Statistical Evidence: Exploring Bases and Implications of the Wells Effect

G. L. Wells (1992) demonstrated that jurors react differently to probabilistic evidence indicating that there is an 80% chance the defendant committed an act than they do to 80% reliable evidence indicating the defendant committed an act. Even though jurors report that the probability of guilt is approximately 80% for both forms, they are much more likely to find against the defendant in the latter than in the former case, a phenomenon the authors term the Wells effect. Three experiments explore a variety of possible explanations for the Wells effect. Results provided the most consistent support for an ease-of-simulation mechanism, which holds that jurors are more reluctant to convict when they can easily simulate a scenario in which the defendant is not guilty.