Effective cutoffs for detecting random, partially random, and nonrandom 350-item MMPI--a short form protocols

The ability of the 350-item short form Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A; Butcher et al., 1992) validity scales to detect random protocols was investigated using samples of 250 nonrandom protocols, 250 half-random protocols, and 250 all-random protocols. As the manual warns, long form cutoffs of the Variable Response Inconsistency scale (VRIN) of 75T and the Infrequency scale (F) of 90T were ineffective in detecting random protocols. Alternative cutoffs for F₁ and the truncated VRIN and F scales were investigated. Short form subscales of VRIN and F were developed to improve detection of partially random protocols. An algorithm using alternative cutoffs for the scales and the new subscales was quite effective, detecting 95% of the all-random protocols, 87% of the half-random protocols, and 98% of the nonrandom protocols. A follow-up cross-validation study was conducted that confirmed the effectiveness of the algorithm.