抄録
I Problem: This is one of a series of studies on the relationship between sociometric choice behavior and interpersonal perception, which has increasingly awoke the interest of many social psychologists. In the previous investigation, we analyzed the relationship in the social situation composed of three persons: the chooser p, the chosen o and the third person q. In this situation three relations are included from p's standpoint: p's attitude toward q (R1), p's attitude toward o (R2), and p's perception of q's attitude toward o (R3). These three relations are considered to be mutually interdependent. For a while, however, let us assume that R3 depends upon R1 and R2, while the latter two are independent of R3 and of each other. As a consequence of the investigation, we concluded that our initial hypothesis derived from Heider's theory should be corrected in one important point.
The conclusion of the previous investigation was based on the data collected in the classrooms which might be considered as “well-structured”, because more than ten months had elapsed since the organization of the classrooms concerned till the date of the data collection. The major purpose of the present investigation, therefore, was to examine if the same relationship would be found in the “un-structured” classrooms as well, and, if not, to see how it would be changed in the course of structuralization of the classroom.
II Method: Subjects of the investigation were ninety-seven pupils in two seventh-grade classrooms in the Faculty School. At the time of the organization of the classrooms, they were almost completely strange one another. They were given relation perception tests six times at intervals of two weeks, the first of which was administered when four weeks had passed by from the organization of the classrooms.
III Results and Discussion: The validity of the method we employed for the purpose of the investigation relies on the assumption that the structuralization of the classroom goes with the time which passed by from the organization of it. The assumption was confirmed by the analysis of several measures.
From the first to the last test, when the total responses were analyzed, the results were very similar to those obtained in the previous investigation. In other words, no differences in result which might be attributed to the different degree of group structuralization were found. Further analysis showed, however, that the relationship between the independent and the dependent variables was related to the time factor in some way. We broke down the total responses into two subgroups: those representing, so to speak, accurate perception and those representing inaccurate one. In the former case the degree to which the hypothesis was confirmed increased progressively as the tests were repeated. In the latter case, on the contrary, it tended rather to decrease.
For these findings a tentative interpretation was presented in terms of different degree of manifestation of q's actual attitude toward o and the different path through which the subjects arrived at the balanced configuration.
