Search results

Hello! I am trying to calculate sample size for a study using paired data. My groups are such that one discordant pair will always be 0. I planned to use McNemar to test for a difference between the groups, however, I am not sure this is valid since one discordant pair will always be zero...