Assumed similarity refers to ascribing similar attributes to the self and others. Because self-other similarity facilitates communion, people who value communion should be prone to assume self-other similarity; but because self-other similarity also evokes obligation, they may be prone to assume similarity only with others with whom they are or would feel comfortable being interconnected. We tested these hypotheses in 5 studies (total N = 1,709). In Study 1, students indicated their political preferences and estimated other students' preferences. In Studies 2-5, students described their personality and the personalities of the following targets: actual or imagined romantic partners in Study 2; ingroup members (students from the respondents' university) and outgroup members (students from a foreign university) in Studies 3-4; and specific liked and disliked others in Study 5. As hypothesized, people with stronger communal values were more likely to assume self-other similarity with liked others, romantic partners, and ingroup members, but not with disliked others and outgroup members. These effects replicated across different cultures (India, Korea, and the United States) and remained significant when controlling for self-esteem, national identification, and attribute desirability. Although people who valued communion tended to depict themselves and liked and ingroup others in relatively normative (typical) ways, which partially explained assumptions of similarity and indicated that those assumptions were to some extent accurate, communal values continued to predict distinctive self-other similarity or "false consensus" even after controlling for the normative prevalence of attributes.

Cultural differences in consistency of responding to questions could be artifacts of cultural differences in acquiescence (an agreeing response style). To highlight this issue, the authors attempt to replicate previous research showing that Koreans are less consistent than Americans are when responding to questions about values. In accord with prior research, the authors find that Koreans are both more acquiescent and less consistent than Americans are. However, the differences between countries in acquiescence partially mediate the differences between countries in consistency. In sum, when using mathematical measures of consistency, researchers must take into account the influence of response styles, and the authors demonstrate two methods for accomplishing that.