Internal states or external stimuli: Observers' attitude judgments and the dissonance-theory-self-persuasion controversy

Abstract

Bem (1965, 1967a) has proposed a self-judgment model of attitude change in an attempt to account for a variety of hypotheses derived from dissonance theory. One derivation from Bem's model is that observer-subjects should be able to discriminate the circumstances controlling the behavior of involved subjects and to estimate accurately the attitudes of involved subjects at the end of the experimental procedure. This derivation has been supported in serveral “interpersonal replications” conducted by Bem. It was hypothesized that these results were obtained because Bem's descriptions of the experimental procedures allowed observer-subjects to perceive systematic differences between conditions in initial attitudes of involved subjects. A series of experiments was performed in an attempt to replicate Bem's results and then to show that observer-subjects cannot accurately predict the results of experiments when perceptions of initial attitudes of the involved subjects are controlled. The results of these experiments indicated that attribution of differential initial attitudes had produced the earlier results and that observer-subjects cannot predict the outcomes of attitude-change experiments when perceptions of initial attitudes are controlled by design.