Please note I acknowledge that Pearson's is typically suitable for normally distributed data, but am I actually incorrect in using Spearman's for normal data? Is there advice that normally distributed data HAS TO be analysed using Pearson's?

My understanding is that Spearman's is a more vigorous statistical test- so if I have proven my correlations already using Spearman's, then this should still be ok and my analysis + correlations are still correct (albeit meeting tougher criteria)... and that the only benefit of using Pearson's is that it might power more significance..which in this case is not necessary since I've already run the tests.

Advice on this would be appreciated!

More Donnchadh O'Sullivan's questions See All
Similar questions and discussions