Computing Probabilities of Type Ⅰ Error and Type Ⅱ Error of Sequential Bayesian Procedures for Testing One-Sided Hypotheses

摘要

In this article, it is shown how to compute, in an approximated way, probabilities of Type Ⅰ error and Type Ⅱ error of sequential Bayesian procedures for testing one-sided null hypotheses. First, some theoretical results are obtained, and then an algorithm is developed for applying these results. The prior predictive density plays a central role in this study.