Besides varying with changes in the positive and negative attributes of chosen and rejected alternatives, dissonance also varies with covariation in either the negative or the positive attributes of the chosen and rejected alternatives. What this means in effect is that the greater the general attractiveness or importance of the choice alternatives, the greater the magnitude of postchoice dissonance.
We have already seen, in the section on negative alternatives, two examples of attempts to manipulate importance ( Mills, Aronson and Robinson, 1959; Rosen, 1961). Them two attempts failed to show the hypothesized effects. Nevertheless, we can assume that importance has some effect if we consider as variations in importance the use of differential rewards or justifications for performing unpleasant tasks in the "forced-compliance" experiments.
In addition to its role in "forced-compliance" experiments, importance has been shown to operate in two "free-choice" studies and one "exposure" study. In all three experiments dissonance was shown to vary directly with the importance of the relevant cognitive elements. Zimbardo ( 1960) in his "exposure" study showed that opinion change