Evidence That the Production of Aversive Consequences Is Not Necessary to Create Cognitive Dissonance

The present authors hypothesized, in contrast to 1 influential revision of cognitive dissonance theory, that the production of aversive consequences is not necessary to create cognitive dissonance and that cognitive dissonance will occur even when aversive consequences are not produced. In Experiment 1, participants drank a pleasant- or unpleasant-tasting beverage and were given high or low choice to write a sentence that said they liked the beverage. Participants threw the paper away once they had written the sentence and then rated how much they liked the beverage. In support of the hypothesis, unpleasant-tasting beverage/high-choice participants liked the beverage more than unpleasant-tasting beverage/low-choice participants. A 2nd experiment replicated this effect, using a different counterattitudinal action and different choice manipulation. By demonstrating that the manipulation of dissonance produced increased physiological arousal, a 3rd experiment suggested that self-perception theory could not alternatively explain the results of Experiments 1 and 2.