Hacker News Reader

Not true.

Actually work sample tests have a .29 correlation to on the job performance, followed closely by structured interview questions.

See: http://www.wired.com/2015/04/hire-like-google/

View on HN

The work I'm referencing specifically called out Google as an example of an overly-complex practice that didn't result in objectively higher quality candidates. Google's overall quality of engineering talent is high because they are quick to push people out the door if they don't meet the high standards. Regardless, it's very difficult to scale Google's hiring practices, and it's overall seen as a growth limiter for them (even by people internally - it can take 3-6 months to hire one candidate).

Reply on news.ycombinator.com