And the common currently practices don't control for that any better. The parent's idea (blind or double-blind screening of candidates) seems like it would, to me, be inherently better if we can figure out how to deal with the logistical complexities it adds. (E.g., how does on-site work? Are there on-sites?) If anything, it would be interesting to weigh against current practices of a few phones screens and a on-site.
(It doesn't seem worse that the world I'm living in, where I'm at the bottom rung doing my damndest to make sure I'm not part of the problem by treating people equally, while at the same time I see some employers offering hiring bonuses for "diversity hires".)
The query in the paper allows the candidate to say "they don't know"; if you force their hand in answering by removing that option, they should be again on equal footing, no?
Additionally, I've also read about mathematical means of recording uncertainty in answers and weighting score with that. (That is, correct answer to which you were more certain net more points, incorrect answers to which you were uncertain lose less points than incorrect answers to which you were more certain.) The end result you're trying to get at is which person is giving better answers and certainties.
> Meritocracy is a lie used by the privileged.
Sure, people can twist things into something they're not, but that doesn't inherently make the original thing wrong or evil.