<i>P</i>技法による人格構造の因子分析 II

The purpose of this study was to analyze the personality traits of three persons separately by means of P-technique factorization in order to examine the problem of generality and uniqueness of personality factors.<br>One male and two female college students, considerably different from one another in the main personality characteristics, were chosen for the experiment and used as the subjects every other day for about one month. Eleven variables were obtained chiefly from objective personality measurements. The centroid method of factor analysis and the orthogonal rotations of axis were applied to each of the correlation matrices based upon the fourteen successive variations of the intra-individual scores.<br>Four significant factors were analyzed from each of the individuals, and the results were interpreted in comparison with one another and with those results which the previous researches had obtained by means of R-technique. First we found that the variations of these three subjects in this experimental period had in general some considerable features in common with the R-technique personality factors. Factors similar to Cattell's “H” and “C” which had previously been approved as the common personality factors were found in each of them. Also a factor which had proved itself to represent the surgency of psychological speed or rhythm was found in the two female subjects. The patterns of loading in those factors which were interpreted in terms of any general factors, however, did not necessarily coincide exactly with one another. Moreover, there were factors which could not be interpreted by any of the general common factor terms and seemed to be quite peculiar to each individual. Accordingly, there were no two factors which were identical both in quality and quantity.<br>Generally speaking, the male subject K was different considerably from others in respect to the whole pattern and also to the supposed ideational upsurgency factor, while the two female subjects, S and N, had somewhat similar patterns but were different from each other in regard to the loadings of the intellectual performance variable.<br>A suggestion could be given as to the theoretical possibility of prediction of individual behaviors by means of the basic formula hypothesized in factor analysis. However, in order to develop this idea of personality prediction, more detailed experimental designs for P-technique factorization should be carried out. Especially, the variables used in this sort of P-analysis should be selected more from observational life-records and behavior-ratings which are really significant to the individual's daily experiences.