就学前幼児におけるロールシャッハ・テストの研究

抄録

Many of the Rorschach studies of children aim to count the frequencies of the Rorschach variables, and lack the longitudinal investigation of them. Besides, there are few reports on the effectiveness, validity and reliability of the test. The purpose of the present study is to throw light upon these points which has been neglected.
The subjects used in the cross-sectional data were 240 children, from 4 to 6 years old, in two private kindergarten and four public nurseries. They were divided into six groups, each of them consisted of 40 children with comparable age and sex.
The results of the present study are summarized as follows:-
(1) The cross-sectional data of the test are obtained from 240 children. The number of responses in this study were fewer than the prior studies, but nearly the same number as those of Ames. The increasing FC with age, and more C in lower age were two of the characteristic findings of this study.
(2) In order to check the cross-sectional data, the longitudinal approach was taken to retest 13 children of 4 years old, selected from the sample two years later. The full agreement between two data was found, validating the cross-sectional data. The tendencies of the increasing FC with age and more C in lower age were proved again.
(3) To examine the effectiveness, the criterion of Klopfer and Margulies was used.“A” pattern, according to the classification made by them, could not be found, but more “B” pattern. This revealed that there are many children who showed the crude relation.
(4) In general, the children who rejected all ten cards were characterized by low intelligence. This kind of subjects in 6 years-of-age group were in the grade of the feebleminded. And the intelligence level of the children who rejected one or more cards was found to be much lower than that of the children who didn't.
(5) The validity was examined from three aspects, i. e., sex difference, intelligence and the rating made by nurses.
As to sex differencies of each variable, boys tended to give more W responses than girls, and the older they were, the more M responses were given. On the contrary, girls tended to give color responses more than boys. There was an agreement between this tendencies and the previous findings.
The correlation between intelligence, in terms of L. Q. and M. A., and the each variable of the Rorschach was calculated. Generally the M responses gave the highest correlation and A% gave a negative correlation.
Ten Subjects were rated individually at 3 nurses' conference with respect to their “behavioral maturity”,“agreeableness in human relation”, and “introversiveness”. These ratings were correlated with the ratings based upon their Rorschach protocols by means of Kendall's tau. The appreciable agreement was given only by 6 years children.
(6) The test-retest reliabilities of the Rorschach determinants were calculated. The reliabilities of six month interval ranged from. 25 to 83, all of them are above the 5% level of confidence except W% The reliabilities beyond two year interval ranged from. 01 to. 52 and R, M, D and A are above 5% level of significance.
The results reported above verified longitudinal regularity and desirable reliability of the Rorschach responses. However, further examinations in the validity and effectiveness of the test are to be needed for subjects who are five or less years of age.