Syndicate content

Add new comment

Submitted by TAKADA SHIN on

I regret to say that the idea presented here is not compatible with the "Neyman-Peason" theory, the basis of discussion with p-value. In the theory, if p>0.05, it implies that you CANNOT deny the null hypothesis: you cannot say anything. Indeed the possibility that the intervention did not work is left, you cannot deduce it scientifically. Of course you can investigate why it is not significant, but then, you are out of "Neyman-Peason" theory and the p-value turns meaningless.