Bayes for beginners? Some pedagogical questions

Ought we to base beginning instruction in statistics for general students on the Bayesian approach to inference? In the long run, this question will be settled by progress (or lack of progress) in persuading users of statistical methods to choose Bayesian methods. This paper is primarily concerned with the pedagogical challenges posed by Bayesian reasoning. It argues, based on research in psychology and education and a comparison of Bayesian and standard reasoning, that Bayesian inference is harder to convey to beginners than the already hard reasoning of standard statistical inference.