2 つの正規母分散間の差を検定するための群逐次検定方式

抄録

In this study we discuss a group sequential test for checking the difference between variances of two normal populations. Specifically, we construct repeated confidence boundaries for the group sequential test. Since it is difficult to determine them satisfying a specified significance level exactly, we determine conservative repeated confidence boundaries. We give simulation results regarding repeated confidence boundaries and power of the test.