ABSTRACT
A graphical-tree representation was used in both the frequency-and the probability-training conditions. If the fact that the frequency representations in the earlier study were graphical accounts for their greater success in training Bayesian inference, then there should be no systematic difference between the effectiveness of the frequency and probability training in the present study. To assess the difference in effectiveness between graphical and nongraphical representational formats, the tree conditions were compared against a rule training using Bayes’ formula. Thus, this study included three types of training: frequency-tree, probability-tree, and rule training.
