Subjects were given questionnaires asking their responses to six statements on public issues. On some of the issues subjects were highly committed to their own views, while other issues were selected for the subjects' lack of knowledge and commitment. Some subjects were asked to respond anonymously, and others were led to believe that their responses would be made public. Each questionnaire contained a bogus distribution of responses from students at the subjects’ university. Each statement of interest was supported by the bogus majority in one questionnaire form and rejected by it in the other form. Several types of conformity were considered. For the low-commitment issues, subjects were influenced in the direction of the bogus consensus. For the high-commitment issues, the bogus consensus had no significant effect. Implications for understanding the effect of polls on public opinion are discussed.