Evaluating insanity: A study of construct validity.

Abstract

In place of the individualistic-subjective approaches to insanity evaluations, a specific criterion-based construct of insanity is proposed and examined. This construct entails the presence of a severe mental disorder which has substantially impaired the individual's cognitive and/or behavioral control. The construct was tested through the utilization of a specifically designed protocol (RCRAS) on 73 insanity evaluations from three forensic settings. Estimates of inter-rater reliability were obtained by independent rating of 25 cases, and yielded satisfactory to exceptionally high reliability coefficients (mean kappa r = 0.77). Results of a discriminant analysis and MANOVA generally confirmed the validity of the construct, demonstrating that insane patients were evaluated as having more severe psychopathology and a greater loss of cognitive and behavioral control.