The present study aims to find out the essential factors representing the aptitude for the various courses of the university. In order to attain to this purpose two kinds of different test batteries were administered to 631 subjects, most of whom were graduate course students of Tokyo University. One battery consists of five sets of test items namely interest, personality, vocational interest, ability and the rating scales for their achievement in high school subjects. And the other consists of rating scales for 144 specialized courses of the university which are designed to get responses under the three kinds of differfent instructions. We call the former predictors and the latter criterions in this study.
The analysis of the data are summarized as follows.
At first it was attempted to reduce the numbers of the predictori and the criteria on the basis of the information quantity defined in terms of the products between variance and the reliability coefficient for each item. By applying principal factor analysis to the selected items, 60 predicter scales and 36 criterion scales are constructed. Next it was attempted to examine whether subjects are adjustedto their courses or not by making use of 15 self rating. scales By means of the factor scores which are yielded by applying factor analysis to 15 scales, 55 subjects are excluded from the following analysis, since they are determined to be mal-adjusted to their courses. On the basis of the data of 576 subj ects which are determined to be adjusted to their courses, distances defined in terms of Euclidian measure among 84 specialized courses are calculated. Then it was decided to apply a kind of cluster analysis to these distances in order to classify 84 groups into a small number of groups. As a result, 12 groups are found to form clusters. They are named as Law, Economics, Literatute, Education, Psychology with Health science, Mathematics with Physics, Engineering, Chemistry, Architecture, Biology, Pharmacy and Medicine.
By applying multiple discriminant analysis to 60 predictor scales and 36 criterion scales respectively, the following four discriminant factors were extracted from each test battery. They are named: (1) natural science vs. humanities,(2) architecture vs. chemistry and medicine,(3) practical business vs. research business,(4) social welfare vs. egocentrism. Then it was presumed that these factors are the essential factors which differentiate the aptitude of the various courses of the university. Then we could conclude that the aptitude of the various courses of the university could be measured to a larger extent in terms of psychological traits including interest and personality dimensions than it has ever been expected.
View full abstract