完全随机设计的方差分析中检验效能的估计

POWER OF ANALYSIS OF VARIANCE IN COMPLETELY RANDOM DESIGN

在实验设计时应保证一定的检验效能；当统计检验结果是阴性时，应估计其检验效能。本文探讨了完全随机设计的方差分析中检验效能的估计方法。结果表明，检验效能与样本含量的大小及样本均数间的差别有关。同时，用医学实例说明其应用。

In this papor,the power of ANOVA in completely random design was deduced in theory. It indicate that the power depends on the distribution function of noncentral F-distribution. At the same time, the medical example was explained.