ABSTRACT
The four approaches discussed in chapter 2 give different explanations for why people do or do not follow probability theory in their judgments. Because the approaches differ in their predictions about whether training is possible and how training about statistical reasoning should look, the results of training studies can be used to discriminate among them. This chapter analyzes all training studies about conjunctive and conditional probabilities, Bayesian inference, and the impact of sample size that could be found in the literature, and examines how their results bear on the four theoretical approaches.1 Because the studies used different designs and dependent measures, an overall metaanalysis could not be conducted, but summaries of subsets of the results are given whenever possible.
