COMMUNITY STANDARDS OF CRIMINAL LIABILITY AND THE INSANITY DEFENSE

Two experiments (N = 71) compare lay standards of insanity to standard s incorporated in American legal codes. In Experiment 1, case vignette s provided only legally relevant information about defendants' degrees of impairment in cognition or in behavioral control. Respondents' jud gments of criminal liability or not guilty by reason of insanity (NGRI ) reflected an exculpatory standard of substantial impairment in both cognition and control. In Experiment 2, case vignettes provided realis tic information about defendants' psychiatric diagnoses; respondents h ad to infer levels of cognitive and control impairment. Results showed that respondents made highly idiosyncratic inferences based on diagno stic categories, but once made, these inferences predicted NGRI judgme nts. Implications of the concordance between laypeople's rules for ass igning NGRI verdicts and the rules used in American legal codes are di scussed.