The predictive accuracy of the Psychopathy Checklist-Revised, Level of Service InventoryRevised, HCR-20, Violence Risk Appraisal Guide, and the Lifestyle Criminality Screening Form were compared in a sample of male offenders. Both correlations and receiver operating characteristics measured the relationship between the instruments and the predictive outcome criteria of institutional misconduct and release failure. Although some instruments performed better across the outcome measures, there were no statistical differences in predictive accuracy among the instruments. C lassification of persons and prediction of behaviors are fundamental goals in forensic settings, and much effort has been spent on pursuing these goals. The attainment of these goals leads to equitable judgments of clients and a more efficient administration of the criminal justice system. Soon after an individual comes into contact with the criminal justice system, there are decision points where an

Recent meta-analysis has demonstrated that attitudes and associates are among the best predictors of antisocial behavior. Despite this finding, there are few psychometrically developed and validated measures of criminal and antisocial attitudes and associates. This study reviews the theoretical and empirical development of the Measures of Criminal Attitudes and Associates (MCAA), which is composed of two parts. Part A is a quantified self-report measure of criminal friends. Part B contains four attitude scales: Violence, Entitlement, Antisocial Intent, and Associates. The MCAA showed reasonable reliability (internal consistency and temporal stability) and appropriate convergent and discriminant validity. Criterion validity was evidenced in the scale's relationship with criminal history variables, and a factor analysis confirmed the four distinct scale domains.

A general model of faking on self-report personality test items is proposed and empirically evaluated. The model predicts that differential test item response latencies should be faster for schemacongruent test answers than for noncongruent responses. Thus, individuals faking good should take relatively longer to endorse socially undesirable test item content than desirable test item content. Conversely, individuals faking bad should endorse socially desirable test item content relatively slower than undesirable test item content. Support for the model was found to generalize across personality inventories and across populations of university students and maximum security prisoners. Conflicting results from previous research are viewed in terms of the model. Further testing of the model's generality and practical relevance is discussed.

No abstract

No abstract