A Population Wide Analysis of MCMI-IV Symptom Validity Scales Administered in the Veterans Affairs Healthcare System

The Millon Clinical Multiaxial Inventory-IV (MCMI-IV) is a psychological assessment tool commonly used in Veteran Affairs (VA) settings. However, no research has examined the MCMI-IV symptom validity scales in the veteran population, where high rates of response bias can occur. This study examined convergent validity of the MCMI-IV scales to the Minnesota Multiphasic Personality Inventory-2-Restructured Form (MMPI-2-RF) validity measures, identified base rates of invalid MCMI-IV validity scores in the veteran population, and explored alternative cutoff scores. All MCMI-IVs administered in the VA (04/2019 - 08/2024, <i>N</i> = 6,466), using data from the Corporate Data Warehouse, were examined. MCMI-IV protocols were correlated with same day administrations of the MMPI-2-RF (<i>n</i> = 1,401) using Spearman's correlations. Disclosure and Debasement positively correlated with overreporting validity scales and negatively correlated with underreporting scales on the MMPI-2-RF, while Desirability showed the opposite pattern (all <i>p</i> < .001). Additionally, the Inconsistency scale positively correlated with MMPI-2-RF non-content validity scales. Less than 1% of MCMI-IV of protocols met the test manual's criteria for invalidity, a significant departure from invalid rates reported on other measures administered to veterans. Diagnostic accuracy analyses suggested new cutoff scores, specifically that overreporting be identified by X Disclosure base rate score (BRS) ≥ 87 and Z Debasement BRS ≥ 84, and underreporting identified by X Disclosure BRS ≤ 49 and Y Desirability BRS ≥ 74. Results suggest that while the MCMI-IV validity indices measure intended constructs, more optimal cutoffs are presented for the veteran population.