Comment by burnished

They show a huge difference in attrition rates after bad interviews in the first two interviews (I might be missing nuance here). They speculate it could be related to other studies in self perception with huge caveats and ample marking of the speculation.