Open torgheh opened 2 years ago
when the sample size in each group is smaller than 5, the total number of permutations you can get is small. For example, if you have 4 cases vs. 4 controls, the total number of permutation is 70, and the smallest p-value you can get is 1/70, also the distribution of permutation p-value can be unstable. so it is doable, but not recommended.
Hi,
Thanks for your amazing work.
I was wondering if you know any method that can reliably produce p-values using the ideas' distance when number of replicates in each group in less than 5? Since I see in your
permanova
method the minimum number of replicates is 5.Thanks.