3 ms·
I agree with the work sample test. In similar threads, some found cheaters, but the followup interview exposed them because they can't speak about the process a
by liquidcool 11y ago
I agree with the work sample test. In similar threads, some found cheaters, but the followup interview exposed them because they can't speak about the process and solution cogently. I imagine that if you look at enough, just reading them will allow you to spot cheaters, and having so many applicants will provide a good selection ethical developers. And a short contract-to-hire or probation period should deal with any who get through.
- crdb 11y agoThe other issue with judging sample code is the sheer amount of time it takes. The two of us took around 3 months (full time, effectively, doing nothing else) to select, interview and make offers to around 10 employees out of a pool of over 100 qualified applicants. You can sort of understand why companies try to develop less time consuming methods, and why experienced developers prefer to hire from their network...
- liquidcool 11y agoThanks for the data on the evaluation time. 9-10 hours per applicant is nontrivial, I agree. If you've worked with someone shoulder to shoulder, I can also see that being as effective. The problem I hear is that depending on the experience and number of the founding team, they can exhaust their network quickly. Peopleware has the idea of auditions, and I thought if you were clear about what you wanted, applicants would be comparable. I'm thinking, "Walk us through a project you did that demonstrates your knowledge of code quality." Or similar, with a clear list of what you're hoping to see.