Date of Award


Document Type

Doctoral Research Project

Degree Name

Doctor of Psychology (PsyD)



First Advisor

Radhika Krishnamurthy, Psy.D., ABAP

Second Advisor

Mariana Juras, Ph.D.

Third Advisor

Marshall Jones, DBA

Fourth Advisor

Robert A. Taylor


Alleged and convicted sex offenders are generally required to undergo forensic evaluations that frequently include a psychological assessment to answer a wide range of psycho-legal questions. A critical component of the psychological assessment of sex offenders is the evaluation of personality and psychopathology. The Minnesota Multiphasic Personality Inventory (MMPI) and its subsequent editions have been the most widely used personality assessment measures in forensic evaluations, including sex offender evaluations, and in research conducted with sex offenders. The previous literature demonstrates that a subgroup of sex offenders often produce test scores reflective of the general population, which is referred to as a “within-normal-limits” (WNL) profile. However, these scores may not accurately represent their psychological adjustment as findings have indicated that sex offenders often engage in minimization of their difficulties. The current study used the Minnesota Multiphasic Personality Inventory-Second Edition-Revised Form (MMPI-2-RF) to identify empirically derived optimal cut scores for the substantive scales (i.e., Higher-Order, Restructured Clinical, Specific Problems, and Personality Psychopathology Five) for sex offenders who produced WNL profiles. Archival data from a sample of N = 122 adult male sex offenders who produced WNL MMPI-2-RF profiles was compared to MMPI-2-RF data from a sample of N =146 adult men from the community. Multivariate analysis (MANOVA), univariate analyses (ANOVA), and Mann-Whitney U test results indicated statistically significant differences between the two samples' mean scores on 28 of the 40 substantive scales, with all but one of the scales significantly lower for the sex offenders. Receiver operating characteristics (ROC) analyses revealed four scales (i.e., RC3, RC7, RC9, COG) produced AUC values at or above 0.70. The optimal cut scores across all MMPI-2-RF substantive scales ranged from 40.5 (FML) to 64.5 (DISC-r). The optimal cut scores were between 40 and 49.5 for 18 scales, 50 to 59.5 for 18 scales, and four scales 60 and 64.5 for four scales. Alternative cut scores at equal 5-point T-score intervals of the ROC-derived cut scores were subsequently evaluated. Results showed that a T score of 40 was optimal for two scales, a T score of 45 for 12 scales, a T score of 50 for 17 scales, and a T score of 55 for nine scales. This study's implications, limitations, and future directions were discussed.