THE MAIN purpose of this investigation was to see if a number of curriculum groups in the health related professions could be differentiated from each other on the basis of personality variables as measured by the Minnesota Multiphasic Personality Inventory (MMPI). Considerable effort is being ex pended at the University of Florida's College of Health Related Professions to find personality, ap titude, interest, and other differences among stu dents in the health related professions (i. e. Occu pational Therapy, Physical Therapy) which will ul timately enable counselors to function more effec tively in the advising of students into and out of these professions. There have been a number of studies where the MMPI was used to determine personality differ ences among college majors. Lough (7) found the MMPI to be of little or no value for detecting differ ences among women students in liberal arts, nurs ing, and teacher training. Norman and Redlo (8) found that certain scales differentiated certain cur ricula from each other for seven curriculum groups of female students. Clark (2) found that profiles for each of a large number of college majors (13 majors for males and 12 majors for females) showed few significant differences from the average college pro f i le. It is of interest to note that the majority of these differences were on the Masculinity-Femi ninity scale, which is interpreted by many investi gators as an interest scale. Harder (5) found no MMPI differences among business, education, and engineering students that were useful is describing these groups. The results of these studies have not been partic ularly encouraging. However, in all of the above studies, the data were handled in a most inefficient manner. The most efficient manner to handle mul tivariate data of this nature is through the use of dis criminant function analysis. Discriminant analysis I involves considering the complete test vector or pro file in a single analysis. Univariate analyses such as those used in the above-mentioned studies do not take into account.the intercorrelations among the measures and consequently do not make full use of the data. For example, it is possible that not one single variable out of a number of variables will sig nificantly differentiate among the groups, while a test of significance involving the complete profile will yield significant differences among these groups. Furthermore, discriminant analysis indicates not only whether the groups differ significantly f r o m one another, but also the practicality of these differ ences, e.g. the effectiveness with which individuals can be predicted to belong to various groups on the basis of their test scores. Discriminant analysis answers this question. Excellent re views of dis criminant analysis are given by TatsuokaandTiede man (10) and Creager (4). Theoretical and computa tional details are treated in detail by Rao (9) and Kendall (6). It is possible that the MMPI would yield useful differences among college majors if the data were handled by multivariate techniques. Consequently, discriminant analysis was utilized in the present in vestigation. The college majors of most interest to the pres ent investigator were Medical Technology (MT), Oc cupational Therapy (OT), and Physical Therapy (PT), since these are the undergraduate educational pro grams in the College of Health Related Professions. However, two other groups were included in the analysis for comparison purposes. These groups are Nursing (N) students and Education (E) majors. Although Ns are not in the College of Health Related Professions, their roles are similar to those of stu dents in these professions. The Es, of course, are the furthest removed from the Health Related P r o I fessions.
Read full abstract