Abstract
People have a tendency to infer attributes corresponding to the behavior of others (correspondence inference). This tendency has also been observed even if the behavior of others is socially restrained (correspondence bias). In this study, we examined the effect that inclusion or omission of the Japanese first person pronoun "I" had on correspondence inference. Study 1 indicated that correspondence inference was more pronounced when participants believed that an essay which included first person pronouns was written by a target person in the no-choice situation, relative to that without first person pronouns. Study 2 found that the greatest influence of correspondence inference occurred when participants who read an essay that included first person pronouns were not instructed to accurately judge the true attitude of a target person. Implications of cultural context-based factors and issues for future research were discussed.
