Thursday, July 23, 2020 [Tweets] [Favorites]

Programming Job Interviews

Chris Parnin and Matt Shipman (via Hacker News, 2, 3):

A new study from North Carolina State University and Microsoft finds that the technical interviews currently used in hiring for many software engineering positions test whether a job candidate has performance anxiety rather than whether the candidate is competent at coding.

[…]

Half of the study participants were given a conventional technical interview, with an interviewer looking on. The other half of the participants were asked to solve their problem on a whiteboard in a private room. The private interviews did not require study participants to explain their solutions aloud, and had no interviewers looking over their shoulders.

[…]

“But the format may also serve as a barrier to entire classes of candidates. For example, in our study, all of the women who took the public interview failed, while all of the women who took the private interview passed.

CobrastanJorji:

I conducted a couple hundred interviews for my first FAANG employer, and I was constantly amazed at the percentage of candidates with years of Microsoft or Facebook experience on the resumes who apparently did not know how to program. I always thought, ‘huh, guess I know why they quit after 3 years, amazing that they all lasted this long.”

Then I interviewed for another company and utterly bombed. It became suddenly clear to me that I had been an idiot. Of course nearly all of those candidates were perfectly good programmers.

Joel Spolsky:

Those 200 resumes you got from Craigslist? Those consist of the one guy who happened to be good, but he’s only applying for a job because his wife wants to be nearer to her family, and the usual floating population of 199 people who apply for every single job and are qualified for none. And now you think you’re being “super selective” but you’re not, it’s just a statistical fallacy.

Somehow this classic hypothetical spread, and people got the idea that the majority of working programmers can’t actually program at all.

See also:

5 Comments

I think it's pretty well-established in the field of organizational psychology that conventional in-person interviews have basically zero predictive power for future job performance. Once every few years, people start talking about this, then they forget it again and go right back to the old way of interviewing people for jobs. The illusion that you personally are just an astonishingly great judge of other people and can establish their skills within minutes is just overpowering all evidence to the contrary.

The exact same applies to performance appraisal.

https://software.rajivprab.com/2019/07/27/hiring-is-broken-and-yours-is-too/

"I can’t keep track of the number of articles I’ve read about hiring in the past few years. Inevitably, they all follow the exact same format. First, they claim that hiring is broken. Next, they describe the hiring practices used by XYZ company. The present an extremely thorough analysis of every possible way in which such a practice can be gamed, or produce bad results. Finally, they wrap up by saying that hiring should instead be done on the basis of ABC criteria, only for someone else to write an essay the following week attacking ABC as a broken hiring method.

At this point, I’ve seen almost every single combination of ABC and XYZ. As a fun little exercise, I thought I’d take some time to compile every one of them."

I can’t keep track of the number of articles I’ve read about hiring in the past few years. Inevitably, they all follow the exact same format. First, they claim that hiring is broken. Next, they describe the hiring practices used by XYZ company. The present an extremely thorough analysis of every possible way in which such a practice can be gamed, or produce bad results. Finally, they wrap up by saying that hiring should instead be done on the basis of ABC criteria, only for someone else to write an essay the following week attacking ABC as a broken hiring method.

See also: software development methodologies. Or task management ones.

The reason new articles on this topic come out on a regular basis is that these things actually *are* broken, we just either don't know how to do them correctly, or we don't listen to people who research these topics, because the solutions they propose don't "feel" right. The former problem mostly applies to software development, the latter mostly to hiring.

Video editors have demo reels they put together over months and years.

Designers have portfolios they put together over months and years.

Developers like to humiliate each other by asking esoteric cs and math questions on the spot.

Should be pretty clear what's going wrong here.

Stay up-to-date by subscribing to the Comments RSS Feed for this post.

Leave a Comment