> N = 15 Can you even judge the sample size separate from effect size? If 15 out of 15 people register an improvement it's a lot more meaningful than 3 out of

> N = 15 Can you even judge the sample size separate from effect size? If 15 out of 15 people register an improvement it's a lot more meaningful than 3 out of a 100 sample, despite the much larger sample size in the latter case.