Abstract
Subjects, who were randomly asigned to the three experimental conditions, watched the actor's achievement behavior (success or failure) on videotape and then made causal attributions for his behavior. I Subjects were instructed to make both observation and causal attributions from actor's point of view (E). II After some task performance, subjects were given the same instruction as I (P). III Subjects were instructed to make both observation and causal attributions from their own point of view (S). The subjects in condition I and II were also asked to make causal attributions from their own point of view at the end of the experiment. The results were as follows: 1. P tended to make more external attribution than S, but less than E. 2. Significant discrepancy was found between the attribution scores obtained from actor's point of view and those from subjects' own point of view in E condition, but not in P condition.
