The Failure to Construct an MMPI-Based Incest Perpetrator Scale

Abstract

The viability of the MMPI item pool for the purpose of constructing a scale to identify and assess the characteristics of incest perpetrators was empirically tested using the contrasted groups or external criterion method. All 566 MMPI items were empirically contrasted through an item selection procedure known as "Validation Generalization," which minimized Type I selection errors. Subjects included 89 male incest perpetrators and 81 male mental health outpatients who had no known history of psychosexual disorder. Out of an initial pool of 81 significant MMPI items, six items were selected that reliably discriminated between the criterion and contrast groups at the .05 level. Only one of these six items was directly indicative of sexual deviance. The obtained number of significant MMPI items was insufficient for inventory scale construction. It was concluded that these findings do not support the utility of the MMPI item pool in making prepetrator identifications in outpatient settings.