中文摘要：

在生物学研究中,为比较两组研究对象的多重结果,在假设总体分布为正态分布的情况下,参数方法如多元的方差分析方法及Bonferroni方法广为应用.然而在多数情况下总体为正态分布的假设并不成立,因而人们自然地提出了更为稳健的非参数检验统计量,如O＇Brien检验统计量,Huang检验统计量,MAX及SAR检验统计量等.本文旨在提出一个更为稳健的检验统计量.

英文摘要：

To compare multiple outcomes between groups in biomedical research,parametric procedure such as multivariate ANOVA or the Bonferroni procedure are used,with the assumption of multivariate normality.Since the assumption of multivariate normality may be false,nonparametric procedures robust against distributions such as MAX test and SAR,are advised naturally.In this article,we provide a new robust test.