In two experiments on choice the durations of attention to the alternatives were measured. In experiment 1 each subject chose one from two pictures; in experiment 2 the choice was one from three pictures. In both experiments the subjects understood that they would acquire the picture that they selected. In each experiment higher and lower conflict conditions were induced by offering subjects a choice between alternatives that had been evaluated either equally or disparately.
In both experiments a significant relationship appeared between duration of attention and preference order with most subjects looking longest at the alternative that was preferred.
In the comparison between conditions this effect was found to be stronger under lower conflict than under higher conflict; this difference reached a significant level in experiment 2. These results are contrary to findings by Gerard (1967), and this matter is discussed. The relevance of the results to other theories is examined.
Inferences were drawn from dissonance theory about re-evaluation effects after decision, and evaluation changes were measured in the experiments. After adjustment for measurement regression, the data failed to reveal a significant chronic re-evaluation effect. Contrary to dissonance theory, the re-evaluation effect was weaker in the three-alternative choice experiment than in the two-alternative choice experiment.