Let's say we have two statistical methods, M1 and M2:
- We state the null hyp: The effect of something is 0
- M1 has less power than M2 (harder to reject a false null, so M1 more often makes type II errors)
Is the following interpretation correct?:
When we reject the null by applying method M1 because of a very tiny p-value, we all else equal end up having stronger evidence for the presence of an effect compared to if we had rejected the null by applying method M2 due to a low p-value.