内田・クレペリン精神作業曲線の因子分析的研究

PO-技法の適用による個人内変異の検討

抄録

Factor analysis has been proposed and used as a method of statistical analysis of several measurements made on one individual repeatedly over a period time.
The purpose of this study was to analyse the intra-individual difference of a single person on the work curve of the Uchida-Kraepelin Psychodiagnostic Test by means of P-O-technique factorization. Furthermore, the purpose was to indicate to what extent the subjects' curve revealed. A unique pattern of factor loadings for each person and to what extent it was stable through the repetition of this test.
Our students sat as the subjects for 40 days in succession. The performance scores of each row were directly used for the raw scores. In P-technique, 30 variables (row 30 of this test) based upon the 40 day to day variation of intra individual differences were correlated, and in O-technique, 40 occasions based upon the 30 row to row variations were correlated. The principal component method of factorization was applied to these matrices of correlations.
The main findings were as follows.
1) The P-factors eliminated seemed to be identified with the factors already obtained by means of R-technique analysis. We might conclude that the greater part of the factor pattern was the same.
2) For the person indicating the higher contribution of factor I to all factor space in P-technique, the considerable stability of his work efficiency could be predicated over the entire time of the series.
3) On factor II and the following factors in P-technique, in spite of a smaller contribution to all factor space, the unique factor loading profiles of each person could be observed. This implied that the combination of these factor profiles made a characteristic feature of the work curve for each person.
4) In O-technique, we could eliminate several factors which indicated the patterns of a multiple work curve. This result implied that the unique curve of one individual was constructed by several different patterns of the work curve. But the higher loadings on factor I over the time of the series indicated that a rather dominant pattern of work curve could existed in one individual, and under the condition, in which factor I occupied the greater part of the contribution to all factor space, cosiderable stability of the work curve in one subject could be predicated.
5) About the factor structure, especially from the point of the factor loadings in factor I and II, we could observe that the work curve appearing in the first test was dissimilar to the curve of a subject having an inner consistency over the time of the series. This implied that, it was impossible to predict the immutability of one particular individual in the results of only the first test.