Identification of random responders on MMPI protocols.

In clinical practice, indicators of random or irrelevant responses on MMPI protocols are typically not examined, and, if obtained, assumed to be the result of either gross psychopathology or attempts to malinger. A sample of 40 computer-generated random profiles and 40 profiles of forensic outpatients were compared on validity scales, clinical scales, and scales specially designed to detect randomness (TR index and Carelessness Scale). Results of discriminant analysis indicated differentiating patterns of responses with random profiles generally elevated. Further, the accuracy of several clinical decisions rules were evaluated. The "F greater than 80 and TR greater than 4" rule was found to have the greatest clinical utility at correctly classifying random responders.