Abstract
The possibility of content bias in the MacAndrew's Alcoholism Scale (MAC) from the MMPI was examined in groups of seizure disorder (n = 35), general medical (n = 25), and psychiatric patients (n = 20). Scores above criteria on the MAC were associated with reported alcohol problems only for the psychiatric patients. Almost 50% of the seizure patients with no such history had scores above criteria on the MAC. This classification inaccuracy is even greater than that reported recently in contemporary normal samples. The results suggest that caution should be used when one is administering the MAC to seizure disorder patients and that further investigation of possible content bias in other MMPI scales or subscales should be pursued.