Conversation

Please retweet: We are currently editing the list of 770 co-signatories to 'Retire statistical significance'! You may still add your name to the list until about Monday 11 March: bit.ly/2tVjv1X, or email to retire.significance2019@gmail.com

Replying to

and

If the model is correctly specified this test is valid. How about we start a movement to end mis-usage? Retiring statistical sig is a bit like making parenting illegal to prevent child abuse.

Replying to

and

Again, it's not against p-values. See my previous comment.

8:22 PM · Mar 9, 2019Twitter for Android

Replying to

and

I cannot see it in this thread. I understand one rationale as the obsession with p-values makes research no longed about research but about chafing sig. stars. Without p-values researchers would have to think & understand, that would be a nice change.

Replying to

and

Nate, you can send

send a one-liner at the above address and have him send the comment so you can have a look. It's a quick read!