Reasonable doubt definition effects on judged guilt: moderation by need for cognition and mediation by changes in required and felt certainty
Journal of Applied Social Psychology
Published online on December 10, 2015
Abstract
Mock‐jurors read a criminal‐trial transcript in which the judge defined “beyond reasonable doubt” with either a stringent (even small doubts require acquittal) or lax (having some doubt does not require acquittal) guilt criterion. Across three experiments, guilty verdict preferences were less likely when the criterion was stringent. A stringent criterion also was associated with higher reported certainty required to vote guilty and lower certainty in the preferred verdict. These two variables mediated the influence of definition on verdict preferences in an analysis that included data from all experiments. Perceived strength of evidence was unaffected by definition. Need for cognition (NC) moderated the definition effect on verdict preferences in all experiments. High‐NC mock‐jurors were strongly influenced by the definition; the influence on low‐NC mock‐jurors was minimal.