data is not normally distributed, non-paired:
Mann-Whitney U test, or named Wilcoxon Rank Sum Test
Ranks all observations from both groups together, calculates the sum of ranks for each group, and uses these sums to assess differences between the distributions.
data is normally distributed, non-paired:
student t test
data is not normally distributed, paired:
Wilcoxon signed-rank test ( don't mix with Wilcoxon rank sum test)
Considers the differences between pairs of observations within the same sample, ranks these differences ignoring the sign, and assesses whether their median differs significantly from zero. 也就是说，先算每个pair的delta，然后rank这些delta Frank Wilcoxon 先在1945提出来的两个non-parametric方法，之后1947, Henry B. Mann 和 Donald R. Whitney 提出了另一个方法（关于U），从数学上来讲和Wilcoxon的signed-rank结果是一样的，所以两个名字都可以。
Use violin plot instead of dot plot/box plot in order to visualize distribution