1 引言在定量资料的假设检验中 ,误用统计方法最常见的原因是忽视资料所对应的设计类型。有些人根本没有这方面的概念 ,只要看到观测结果是定量的资料 ,就认准一定要用ｔ检验。仅从设计类型角度看 ,适合选用ｔ检验的设计类型只有 3种 ,即单组设计、配对设计和成组设计[1] 。这 1 Introduction In the hypothesis testing of quantitative data, the most common reason for the misuse of statistical methods is to ignore the type of design that corresponds to the data. Some people simply do not have the concept of this area, as long as the observation results are quantitative information, we must look for the t-test must be used. Only from the design type point of view, there are only three types of designs that are suitable for the t-test, namely single-group design, paired design and group design [1]. This