Abstract
In the context of linear mixed effects models, we derive an asymptotically efficient group sequential testing procedure fitted with a data dependent allocation rule for comparing the effects of two treatments. The allocation is designed so as to favor procedures that stop early while allocating less patients to the inferior treatment. A simulation study is used to compare the performance of the proposed procedure to those of four standard procedures.
Acknowledgment
This research was supported by a grant from the Swiss National Research Foundation (No. 21-40447.94).