The psychometric assessment of alcoholism in forensic groups: the MacAndrew scale and response bias.

Although the MacAndrew Alcoholism scale is the most widely used Minnesota Multiphasic Personality Inventory (MMPI) measure of vulnerability to alcohol abuse, its accuracy has not been studied in patients intrinsically motivated to exaggerate or minimize psychopathology. We examined the usefulness of the MAC in predicting alcohol abuse in a forensic clinical sample. Results indicate the MAC (a) was not more effective than direct inquiry in this group, (b) scores were correlated negatively with minimization and positively with exaggeration for subjects with histories of alcohol abuse, (c) offered advantages over direct inquiry both in screening for alcohol history (sensitivity) and in confirming it (specificity), and (d) scores were only moderately more accurate in valid than in minimized or exaggerated MMPI protocols. The results suggest that clinicians should use the MAC cautiously, particularly when they suspect motivation to minimize psychopathology.