田间试验中,经常是几个处理或几个品种之间相互进行比较。对这些资料进行方差分析的F测验,只能笼统地说明各处理间有无显著差异,并未具体指出某几个处理间平均数的差异显著与否。要测验各个处理间的具体差异情况,必须进一步进行平均数间差异显著性测验。最早的一种方法:是L.S.D.法(最小显著差数法)。这种方法由费雪(R.A.Fisher)提出,自三十年代沿用至今。它是以t测验为基础的。但是t测验法只适用于测验两个平均数间的差异显著性。多个平均 In field trials, it is often the case that several treatments or several varieties are compared with each other. The F-test of variance analysis of these data can only provide a general description of whether there is any significant difference among the treatments, and does not specifically indicate whether the difference between several treatments is significant or not. To test the specific differences between the various processing conditions, the average number of differences between the significance test must be further. One of the earliest methods: the L.S.D. method (least significant difference method). This method was proposed by R.A. Fisher and has been in use since the 1930s. It is based on t-test. However, the t test applies only to test the significance of the difference between two averages. Multiple averages