Multi-linear regression analyses.

Evaluation of the three entropy measurements (within-subject, inter-subject and pooled) with respect to average response times per subject per task . Models include the average response time of each subject during the experiment and a constant . Beta values correspond to the standardized coefficients. reflects the fraction of variance of explained by the model. Results indicate that, when is fixed, is the best descriptor in order to explain . This model containing the term had also the highest . Even when analyzing only the consistent decisions per subject per task, had a significant predictive capacity and the model was able to explain about half of the variability of .