>>Yes, it's fundamentally unfair to those interviewees who don't cheat. And the process is broken. However, the conclusion this person reaches - we must conduct these broken interviews in person so the cheaters can't cheat - is the wrong one.
Correct! Making a harder-to-cheat "closed-book-test" isn't the answer.
A better answer would be to make it "open-book" — allow and expect interviewees to have their full libraries and reference sites open and ready to go. This is actually how they will be doing their work (we hope!). Then judge on efficient lookups AND insightful comments on how to build upon the found knowledge.
Or, based on the above sample recognize that the process is really totally broken, that almost anyone coming in will work out well (the cheaters clearly lacked the skill or confidence of OP, yet did fine in the actual job), and either fix the criteria (obviously difficult based on the last decade of HN discussions on the hiring sh^tshows), or just cut wasting resources on extensive interviews and take the first N applicants and be done with it.
Or how about, you know, stop pretending we can evaluate someone's "intelligence" or "skill level" as if it's a one-dimensional factor that you have superhuman insight into measuring.
I'm satisfied, in an interview, if I feel the person is interested in the topic, is able to admit that they don't know something, and seems like they'd be pleasant (although i'd settle for "not difficult") to work with.
The breadth and depth of their knowledge is secondary and it's always a gamble because people excel at putting up very convincing façades of whatever it is that they think they're being evaluated for.
People have different things to offer, and you really don't find out until at least 6-12 months into their employment what they really excel at, you have no chance in 30-60 minute (or whatever) interview.
The best you can hope for is to rule out the obvious sociopaths.
>The breadth and depth of their knowledge is secondary
That's what the parent comment is saying. Open-book instead of closed-book, just have the candidate walk you through figuring out what they don't know, listen to them think aloud, and evaluate that.
Correct! Making a harder-to-cheat "closed-book-test" isn't the answer.
A better answer would be to make it "open-book" — allow and expect interviewees to have their full libraries and reference sites open and ready to go. This is actually how they will be doing their work (we hope!). Then judge on efficient lookups AND insightful comments on how to build upon the found knowledge.
Or, based on the above sample recognize that the process is really totally broken, that almost anyone coming in will work out well (the cheaters clearly lacked the skill or confidence of OP, yet did fine in the actual job), and either fix the criteria (obviously difficult based on the last decade of HN discussions on the hiring sh^tshows), or just cut wasting resources on extensive interviews and take the first N applicants and be done with it.