Two-Stage Equivalence Tests That Control Both Size and Power

Abstract

We consider equivalence tests that control both size and power for a difference of means from , i = 1,2. We give equivalence test procedures in two-stage sampling for both the cases and . The test procedures control both size and power with second-order accuracy. We also consider some related problems in clinical trials. Two examples are given to illustrate how the methodologies should be used.

Keywords:

• Clinical trials
• Sample size
• Second-order analysis
• Testing for equivalence
• Two-stage sampling

Subject Classification:

• 62L10
• 62L12

ACKNOWLEDGMENT

I would like to express my deepest gratitude to Professor Makoto Aoshima, University of Tsukuba, Japan, for his invaluable suggestions and guidance, and his continuing encouragement throughout this research.