Completely Randomized Designs for Comparing Dual with Single Treatments
研究了排除一个处理组合的n×m完全随机因子实验，给出了A最优分配的充要条件，并推导出一系列等重复A最优设计。
SUMMARY This paper considers n × m completely randomized factorial experiments in which one treatment combination, labelled 00, is excluded. The aim of the experiments is to compare the effects of having both factors at non-zero labelled levels with the effects of having only one factor at a non-zero labelled level. A necessary and sufficient condition is given for an A-optimal allocation of units to treatments, and a series of equireplicate, A-optimal designs is derived.