[Analysis of non-optimal behavior in a dichotomous choice situation].

Proteau and Girouard (1987) have recently proposed a model to explain the decisional process that takes place in a two-choice reaction time situation. Although the model was well supported by the experimental data reported by these authors, a few unexplained results were observed and hypotheses were suggested to account for them. The goal of the present experiment was, first, to replicate the results presented by Proteau and Girouard (1987) to support their model. Our second goal was to determine if the unexpected results reported by these authors can be linked to some particular conditions occurring in the normal course of an experiment. The results obtained in the present experiment totally supported the proposed model. Furthermore, a detailed analysis of the circumstances under which unexpected results occurred did not permit the identification of any particular conditions favoring their occurrence. In conclusion, the proposed model, although reliable for about 95% of the trials, is still imperfect.