AND

Two studies examined certain discrepancies which have been con-sidered important evidence in support of interpersonal comparison (value-adherence) explanations of group induced shifts in choice. These are (a) the differences between a person’s own choice and the choice he predicts others would make and (b) the differ-ence between the former and the choice he admires. Findings from the first study indicate that own choices are more extreme than those a person predicts others would make because he is more certain and confident about the former than the latter, not because he wishes to appear to outdo others as interpersonal comparison theories of choice-shift effect would have it. The second study strongly suggests that extreme choices are admired not because they display maximal adherence to a social ideal but because they imply that the person’s solution to a problem involving choice is well-founded, that he has persuasive reasons for the choice. On the whole the evidence bodes well for explanations of choice-shift effects based on persuasive argumentation and poorly for those relying on interpersonal comparison processes. Despite prolonged sermons (e.g., Moscovici & Lecuyer, 1972) and brief laments (e.g., Smith, 1972) to the contrary, research on group induced revisions in individual choice-the ill-named “risky shift”-has in fact progressed to where critical experimentation seems possible. The issues at hand in this area have to do with the necessary and sufficient condition