Abstract
In this study, we propose a group sequential procedure that allows the change of necessary sample size at intermediary stage in sequential test. In the procedure, we formulate the conditional power to judge the necessity of the change of sample size in decision rules. Furthermore, we present an integral formula of the power of the test and show how to change the necessary sample size by using the power of the test. In simulation studies, we investigate the characteristics of the change of sample size and the pattern of decision across all stages based on generated normal random numbers.
Acknowledgments
The authors are deeply grateful to the referees for their helpful advice and useful comments on the article. I express my appreciation to Associate Professor Yoshiro Yamamoto and Toshiharu Ikeya of Tokai University for their cooperation in calculations for simulation studies. Special thanks are due to the editor for his constructive comments and suggestions.