Abstract
The purpose of this study was to explore individual differences in tendency to restore balanced state (Heider, 1946, 1958), as applied to both real and fictitious p-o-q triads. According to the previous investigations, balance hypothesis has been substantiated only when sign of p's relation toward o is positive. Therefore, the present study dealt with triads where p liked o.
One hundred and fifty-three girls' highschool students served as subjects (Ss). In real situations, they were, first, asked to rate each classmate on a three-point-scale: “Like”, “Neutral”, and “Dislike” . Data concerning p's sentiments toward o and q were thus collected. Ss were given, a week later, the “relation-perception test” which had been prepared for each particular S, and where Ss were requested to guess o (liked person) 's sentiment toward q (liked or disliked persons) on a five-point-scale. Individual score for tendency to balanced state was determined from each S's responses to the test. In fictitious situations, o and q were hypothetical, instead of real, persons. Ss were asked to assess o's most probable sentiment toward q in general.
Ss were broken up into three groups according to their scores in cognitive complexity test (ERCQ) after Crockett (1965).
The results obtained were as follows: (1) In fictitious situations, Ss low in complexity were more prone to rely on the balance schema than those intermediate; (2) in real situations, on the contrary, Ss high in complexity were more prone to rely on the schema than those intermediate or low; and (3) it was suggested that the relationships between cognitive complexity and tendency to balanced state were curvilinear.
From these results, some implications of the balance schema were discussed.
