Does the Neurological Damage from a Traumatic Brain Injury Invalidate Measures of Psychopathology Such as the MMPI-2?

Abstract

The Gass correction model removes 14 items from the MMPI-2 which refer to neurological damage when assessing a person suffering a traumatic brain injury (TBI). Previous analysis supported ongoing inclusion of the five correction items from MMPI-2 Hs scale in individuals who suffered a TBI. This study investigates the remaining nine correction model items on the Hy and Sc scales. A sample of patients with a TBI (n = 242), and the MMPI-2 normative sample (n = 2,600). This is a cross-sectional study testing measurement invariance of two MMPI-2 scales (Hy and Sc) between the TBI and normative samples. Factor models for each scale were defined and assessed across multiple samples. The Hy scale generated a four-factor model (CFI range .913-.935, RMSEA range .025-.027). The Sc scale generated a five-factor model (CFI range .911-.962, RMSEA range .013-.018). Both models met the criteria of partial measurement invariance. Sensitivity and specificity analysis supported retaining all correction items when assessing persons with a TBI. These results suggest continued inclusion of all 14 Gass correction model items when assessing a person with a TBI.