I have noticed that rank is calculated by counting the significant p-values per interaction pair (per row) and dividing with the total number of cluster-cluster comparisons. In that case, the rank value should be no greater than 1. However, I found many results with a rank value over 1. So, why is it?
Hi,
I have noticed that
rankis calculated by counting the significant p-values per interaction pair (per row) and dividing with the total number of cluster-cluster comparisons. In that case, therankvalue should be no greater than 1. However, I found many results with arankvalue over 1. So, why is it?