A must-watch talk by Richard Morey here. When you claim you're using a flawed method to raise awareness, "you're no longer in it for the truth but you're playing the game."
"You should be skeptical of basically all meta-analytic forensic methods... This includes p-curving."
Follow
I think that’s a fair summary:
Looking for at the “reported p-values around 0.05” is a flawed process because:
1) It assumes different scientific process than really happen
2) it has especially and tendency to flag slightly biased reports that are not problematic
3) there is too much trust in simulation
Do you agree?
I’m trying to make sure I understand.
@plenartowicz Largely on point, especially the first one. Also see this blog post if you haven't. https://twitter.com/jbakcoleman/status/1634333272120471552?t=yIJ456jwrcjfSh8I6qN_Xg&s=19