The effect of interaction anticipation upon the extremity of trait ratings MICHAEL H. BOND* Stanford University DONALD G. DUTTONt University of British Columbia ABSTRACT This study examines the effect of interaction anticipation upon the perception of a future partner. Subjects were led to anticipate interaction with a stimulus person for either 3 or 30 minutes. Control subjects anticipated either no interaction or interaction with a different person. A process of "commitment reluctance" was discovered: sub- jects in experimental groups made less extreme ratings across warm-cold dimensions of stimulus persons who had been described either as "warm" or as "cold." These rat- ings were not, however, related to the length of the expected interaction. This experiment was designed to discover whether the expectation of future interaction with an individual affects the assessment of information about that person. The study is largely exploratory, as there is little related work in the literature. One exception is a study by Mirels and Mills (1964). To test one of their hypotheses, these experimenters asked two groups to rate stimulus ma- terials about a person who was presented as unpleasant. One group had been informed that this person was to be their future partner in a problem- solving task; the other group was told that this person was not to be their partner in that task. The former group rated their partner as less un- pleasant than did the latter, confirming a prediction derived from Fes- tinger's theory of cognitive dissonance (1957). One problem in the interpretation of their study, however, emerges from a recent study by Manes (1969). She asked subjects to memorize a list of contradictory information about another person, up to a low criterion of retention. Subjects in this "low-learn" condition showed more polarized judgments of the material than did subjects who learned the list more thoroughly. It is possible, however, that such polarization could also occur with consistent information, as in the MircLs and Mills (1964) study. It could be argued that subjects who anticipate interaction with the person described will learn information about that person more thoroughly than * Now at Department of Sociology, Kwansei Cakuin University, Nishinomiya, Japan. f Requests for reprints should l>e sent to Donald C. Dutton, Department of Psy- chology, University of British Columbia, Vancouver 8, British Columbia, Canada. CANAD. J. BEIIAV. SCI./REV. CANAU. SCI. COMI\, 5(3), 1973