An evaluation of the MMPI-2 using South African pre-trial forensic patients : prediction of criminal responsibility and assessment of personality characteristics

by Du Toit, Emile

Abstract (Summary)
This study examines the utility of the MMPI-2 in predicting responsibility in pre-trial forensic patients, using a post hoc sample of 94 offenders from Sterkfontein Psychiatric Hospital in Gauteng. Firstly, the overall characteristics of the pre-trial forensic patients are discussed, following an analysis of demographic, clinical, criminal and MMPI-2 pre-trial data, as well as an overview of the Megargee typological classification of offenders. The sample is classified into Criminally Responsible (CR), Diminished Criminal Responsibility (DCR) and Not Criminally Responsible (NCR), and the CR and DCR groups are collapsed (CR/DCR) for many of the analyses when comparing them to the NCR group. Secondly, the variance of variables with responsibility is discussed, after examining one-way ANOVA’s of demographic, clinical, criminal and MMPI-2 variables, as well as an overview of high point pairs. Thirdly, discriminant analyses were conducted of demographic, clinical and MMPI-2 variables. When comparing the collapsed CR/DCR group to the NCR group, psychiatric diagnosis, presence of psychosis, the MMPI-2 Pa and Es scales, as well as race and substance abuse each had unique predictive power and created a substantial discriminative equation (F (6,70) = 45.732, p <0.0005) with a successful prediction rate of 96%. Using only MMPI-2 variables to predict responsibility showed significant unique contributions for the Pa, Es, MAC-R and Mf scales, with the BIZ scale not quite significant, and a fairly significant overall discriminant equation (F (5,73) = 6.474, p < 0.0005), with an overall successful prediction rate of 82%, with the MMPI-2 variables adding an additional 3% to the predictive power of the demographic and clinical variables. Similarly, when examining the more complex 3 group responsibility classification of CR, DCR and NCR, it was found that the demographic, clinical and MMPI-2 variables of psychiatric diagnosis, psychosis, race, substance abuse, and the Pa, Es and Ma scales all had significant contributions to a powerful discriminant analysis (F (14, 136) = 19.758, p < 0.0005) that was capable of correctly reclassifying almost 95% of the sample, and the MMPI-2 variables providing an increase in predictive power of 8%. Differences in responsible and not responsible pre-trial forensic patients are discussed, as well as the role of the MMPI-2 in assessing these differences, and the fact that it is highly likely that it adds more to the forensic assessment of responsibility than a 3% (CR/DCR versus NCR) or 8% (CR versus DCR versus NCR) increase in predictive power. Limitations of the study are discussed, together with recommendations for future research with the MMPI-2 for assessment of criminal responsibility. The suggestion is made that the MMPI-2 can become a valuable tool in South African forensic settings, not only in the assessment of responsibility and malingering, but also in the placement, management, follow-up and treatment of offenders, to maximize the limited resources in South Africa allocated for the rehabilitation of offenders, and minimize the risk of recidivism or rehospitalization.
Bibliographical Information:


School:Rhodes University

School Location:South Africa

Source Type:Master's Thesis



Date of Publication:01/01/2004

© 2009 All Rights Reserved.