Revised neurobehavioral scales of the MMPI: sensitivity and specificity in traumatic brain injury

The ability of 23 previously identified Minnesota Multiphasic Personality Inventory (MMPI) "neurologic content" items to distinguish between individuals with traumatic brain injury (TBI; n = 32) or spinal cord injury (SCI; n = 17) was examined. Principal-components analysis of the 23 items revealed three conceptually coherent, nonoverlapping, and uncorrelated factors (Cognitive, Somatic, Inactivity) that together accounted for 44% of the total variance. Coefficients of internal consistency for the factors were in the moderate to high range. Together, the factors were named the Revised Neurobehavioral Scales of the MMPI. The group with TBI scored significantly higher on the Cognitive scale and significantly lower on the Inactivity scale than the group with SCI (with or without depression as a covariate). The Glasgow Coma Scale correlated significantly and negatively with the Cognitive scale in the group with TBI. Discriminant function analysis revealed that together the scales correctly classified individuals with sensitivity and a positive predictive value (with respect to TBI) of 87% and 81%, respectively. Specificity and a negative predictive value (with respect to SCI) were 68% and 76%, respectively. The overall rate of correct classification of individual cases was 80% (with or without depression in the analysis). The Cognitive scale alone correctly classified individuals in the group with TBI with a positive predictive value of 84%. Findings are discussed in terms of the discriminative validity and potential utility of TBI-related MMPI items, as well as the issue of "neurocorrection" of the MMPI (or MMPI-2) in verified cases of TBI.