ABSTRACT
We present a flexible group sequential procedure for comparing several treatments to a control. Though longitudinal data corresponding to a two stage mixed effects model are considered, ranges of application include any process with independent increments. The procedure allows the experimenter to drop the inferior treatments from the trial as soon as they are detected. It control strongly the familywise error rate. We also discuss a new error spending function (ESF) and study the performance of the procedure using various ESFs and time scales. Finally, the procedure is illustrated on a real example and implementation considerations are discussed.
ACKNOWLEDGMENT
The authors wish to thank a referee for his insightful remarks which have substantially improved the paper. Both the first and second authors were partially supported by a grant from the Swiss National Research Foundation (No. 21-40447.94).