VALIDITIES AND INTERCORRELATIONS OF MMPI SUB- SCALES PREDICTIVE OF COLLEGE ACHIEVEMENT1

THE value of a personality questionnaire which is uncorrelated with measures of intellectual ability but is correlated with school achievement is obvious. Presumably, such an instrument would ac-count for a portion of the variation in school performance unac-counted for by measures of scholastic aptitude. In a multiple regres-sion equation, the noncognitive scale would materially increase the predictive efficiency of college entrance examinations. There have been many efforts to devise and validate such an instrument. For example, Kleinmuntz, (1962) lists eleven scales developed from the pool of items in the MMPI which were developed to predict aca-demic success among college populations. One such scale, containing 26 MMPI items, was constructed by Altus (1948) and was found to be correlated.39 with grade-point average and.21 (nonsignificant) with a measure of intelligence, the Altus Measure of Verbal Apti-tude. The purpose of the present paper is to determine the intercorrela-tions among MMPI subscales selected from among those described by Kleinmuntz. At the same time, the study will attempt to deter-mine the relationship of each subscale to a measure of scholastic achievement (grade-point average) and to scholastic aptitude (American College Test, or ACT). This study can be viewed as a combination of predictive validity and postdictive validity, since the administration of the noncognitive predictors occurs between the 1 These data were gathered during the author’s tenure at New Mexico State University and the study was facilitated by a grant from the New Mexic