Dissociating judgment from response processes in statement verification: the effects of experience on each component.

This research used the stochastic judgment model of statement verification to demonstrate a dissociation between judgment and response processes and investigated hypotheses about the effects of practice on each component. Data from respondents judging statements as true or false under various payoffs supported the dissociation and the following conclusions: Their ability to discriminate true from false depended on knowledge domain but not on payoffs. Experience with the domain did not improve this ability but did decrease the trial-by-trial confidence variability associated with memory search. Practice in a different domain had no such effect. Response criteria depended only on payoffs and experience. Criterion variability decreased with cumulative practice over domains. Most respondents had a bias to say "true" under symmetric payoffs, which did not dissipate with experience. Theoretical implications of the results are discussed.