spurious(?) effect of gender and relationship type on the similarity of perceived attitudes

The present survey investigated the effect of gender and relationship type on the similarity of perceived attitudes. 888 Chinese university students completed the Social Relations Questionnaire, nominating those people with whom they had recently discussed important matters and describing their own and their nominees' attitudes on 16 life values. Perceived attitudes were more similar among females and among friends than among males or among relatives and other people. In all cases, respondents' own attitudes were more similar to their perceptions of their own partners' attitudes than to the perceived attitudes of the partners of other randomly selected respondents. However, the effect of gender and relationship type on the similarity of perceived attitudes turned out to be spurious because the same relationships were found to be true for the randomly selected pairs. I argue that the perceptions of one's partners' attitudes may be relatively accurate, since they mirror the actual patterns existing in society, and that social influence may be more important than friendship choice as a cause of similarity, since there is just as much of a similarity increment for relatives as for friends.