Tests for the Ratio of Two Within-Subject Variances in a Parallel Design

Description:

A company has developed a generic drug for treating rheumatism, and wants to show that it has a different within-subject variance of joint-fluid measurements than the standard drug. A parallel-group design with replicates will be used. Between 2 and 5 replicates for each subject will be considered.

Company researchers wish to determine the number of subjects needed to achieve 90% power, for assumed variance ratio values between 1.3 and 2.0. They plan the two group sample sizes to be equal.

The first part of the report shows the needed sample size for each parameter combination.

The sample size chart gives a visual comparison for the varying numbers of replicate counts and assumed variance ratios. For example, there is a dramatic sample size benefit if the number of replicates is 3 rather than 2, but not much benefit from going to 5 replicates from 4.