对于两个以上独立样本的平均数差异的显著性检验，如果并不符合参数检验的前提，应该使用克-瓦式单向方差分析
For significance testing of differences in means of two or more independent samples, if the premise of parametric testing is not met, one-way ANOVA with Kruskal-Wallis should be used.

克-瓦式单向方差是将所有样本的数据合并在一起，按照从小到大的顺序编秩次，再计算各样本的秩次和\\One-way ANOVA with Kruskal-Wallis is to combine the data of all samples, organize the ranks in order from smallest to largest, and then calculate the sum of the ranks of the samples.

克-瓦式单向方差分析的统计值H的值
The value of the statistic H for the Kruskal-Wallis' one-way ANOVA

N是所有样本容量之和，n是各个样本的样本容量，R是各组样本数据的秩次和\\N is the sum of all sample sizes, n is the sample size of each sample, and R is the rank sum of each set of sample data

当样本容量或样本数目比较大的时候，统计量H接近自由度k-1的χ²分布，因此可以通过查χ²分布表来得到相应的H的临界值\\When the sample size or number of samples is relatively large, the statistic H is close to the χ² distribution with degree of freedom k-1, so the corresponding critical value of H can be obtained by checking the table of χ² distribution