Glenn D Walters1. 1. Federal Correctional Institution, Schuylkill, PA, USA. gwalters@bop.gov
Abstract
INTRODUCTION: Several different factor models have been proposed for the Psychological Inventory of Criminal Thinking Styles (PICTS), an 80-item self-report measure designed to assess thinking styles believed to maintain a criminal lifestyle. AIMS: To perform further tests on the factor structure with two new samples, and exploit their divergent characteristics to test for generalizability of the result. METHOD: A series of factor analyses were performed on the PICTS, completed by two groups of offenders: 519 male prisoners and 227 female prisoners, to determine whether a one-factor, two-factor, four-factor or eight-factor model best fitted the data. RESULTS: The two-factor model was superior to the one-factor model, and the four- factor model was superior to both. An eight-factor model, the factors being identical to the original eight PICTS thinking style scales (Mollification, Cutoff, Entitlement, Power Orientation, Sentimentality, Superoptimism, Cognitive Indolence, Discontinuity), achieved a significantly better fit than any of the other models in the male sample; the four and eight factor models were not significantly different among the women. DISCUSSION: Modest to moderate fit was attained with the two-, four- and eight-factor models when the root mean square error of approximation (RMSEA) was utilized but not when the goodness-of-fit index (GFI) was applied. It is concluded that the factor structure of the PICTS consists of two primary or major factors (problem avoidance, self-assertion/deception) and two secondary or minor factors (interpersonal hostility, denial of harm). The four-factor model seems likely to be generalizable across divergent offender samples. Copyright 2005 John Wiley & Sons, Ltd.
INTRODUCTION: Several different factor models have been proposed for the Psychological Inventory of Criminal Thinking Styles (PICTS), an 80-item self-report measure designed to assess thinking styles believed to maintain a criminal lifestyle. AIMS: To perform further tests on the factor structure with two new samples, and exploit their divergent characteristics to test for generalizability of the result. METHOD: A series of factor analyses were performed on the PICTS, completed by two groups of offenders: 519 male prisoners and 227 female prisoners, to determine whether a one-factor, two-factor, four-factor or eight-factor model best fitted the data. RESULTS: The two-factor model was superior to the one-factor model, and the four- factor model was superior to both. An eight-factor model, the factors being identical to the original eight PICTS thinking style scales (Mollification, Cutoff, Entitlement, Power Orientation, Sentimentality, Superoptimism, Cognitive Indolence, Discontinuity), achieved a significantly better fit than any of the other models in the male sample; the four and eight factor models were not significantly different among the women. DISCUSSION: Modest to moderate fit was attained with the two-, four- and eight-factor models when the root mean square error of approximation (RMSEA) was utilized but not when the goodness-of-fit index (GFI) was applied. It is concluded that the factor structure of the PICTS consists of two primary or major factors (problem avoidance, self-assertion/deception) and two secondary or minor factors (interpersonal hostility, denial of harm). The four-factor model seems likely to be generalizable across divergent offender samples. Copyright 2005 John Wiley & Sons, Ltd.
Authors: David J Disabato; Johanna B Folk; John Wilson; Sharen Barboza; Jordan Daylor; June Tangney Journal: J Psychopathol Behav Assess Date: 2015-12-19