Null - Exactly right. There is no large N problem with p-values there is only an incorrect null problem. If you're not interested in small effect sizes then change the null to be: H_0: mu <= c for some interesting c rather than taking c = 0. https://t.co/Wo1qF315ko

4:17am PST - February 24th, 2021

Exactly right. There is no large N problem with p-values there is only an incorrect null problem. If you're not interested in small effect sizes then change the null to be: H_0: mu <= c for some interesting c rather than taking c = 0. https://t.co/Wo1qF315ko

0 retweets 0 likes -

Sam Davenport (@BrainStatsSam)