Preference of dissonance reduction modes as a function of their order, familiarity and reversibility

Abstract

Abstract Two central hypotheses of the original version of the theory of cognitive dissonance (Festinger, 1957) (1) that dissonance is to be conceived of as a primary drive and (2) that in order to reduce dissonance less resistant cognitions will be changed more than highly resistant ones led to hypotheses which were confirmed by two experiments. In Experiment I, a 2 × 3 factorial design, order and familiarity of dissonance reduction modes were manipulated. After receiving a dissonant information the subjects were offered a relatively low and a relatively high resistant cognition for dissonance reduction, each being placed first (series I, high‐low resistance) or last (series II, low‐high resistance) respectively. Subjects did (known) or did not (unknown) read these modes before reacting to them. Under the unknown condition dissonance will be reduced more with a specific mode if it is placed first than last. With known reduction modes the order of presentation does not have an effect. Under series I condition the first‐placed, higher‐resistant cognition will be changed more in the unknown condition than in the known condition. Under the series II condition the first‐placed, lower‐resistant cognition will be changed equally in the known and unknown condition. In short, the higher‐resistant cognition will be changed more, only when it is placed first and when the following modes are not known. Two explanations for these results are possible: (1) The more dissonance is reduced by changing a more or a less resistant congnition, the less further reduction is necessary; (2) dissonance will be reduced in an internally consistent way. Experiment II excluded the first explanation. Subjects were allowed to revise their original way of reducing dissonance. First, subjects in one condition received series I unknown and subjects in the other condition received series II unknown. Reacting to the dissonance reduction modes the second time, there was more revision when the high‐resistant congition was placed first (series 1 revision) than when placed last (series II revision). These results support the hypothesis that dissonance reduction by changing a less‐resistant congnition more and changing a high‐resistant one less is preferred. Implications of the results of the two experments for the stability of dissonance reduction, the method and interpretation of dissonance experiments are discussed.