Search

Two experiments indicated a decreased effect of disconfirmed performance expectancies (a) when nonvalued performance scores were given to subjects, and (b) when the modeling effects of the experimenter were reduced. According to the dependence interpretation advanced, the traditional use of valued performance scores increased the subjects' dependencies on the experimenter and, consequently, the likelihood of their imitating a change in his behavior; that is, disconfirmed performance expectancies may lead to dependence reduction and not necessarily to dissonance reduction. [ABSTRACT FROM AUTHOR]