This OC isn't grinding an axe. I may be wrong, but it seems this paper's contribution is using analytical techniques (whatever "correspondence testing" is), wi

This OC isn't grinding an axe. I may be wrong, but it seems this paper's contribution is using analytical techniques (whatever "correspondence testing" is), with the customary roles reversed, to better calibrate those techniques. If so, I think that's genius. So many times I've found myself facing a (ethical) problem with nothing more than noob wild-ass-guesses on how to solve it. Perhaps better understanding hiring discrimination in all contexts will help divine context-free rules. Neutral strategies we can all agree to use. That'd be awesome. Because right now the only notion I have for tackling bias is using lotteries (eg sortition). And that's a hard sell even for me.