3 ms·
A thought-provoking post. I really like the specific details the author provides about the interview set-up and the results observed among candidates at his com
by tokenadult 14y ago
A thought-provoking post. I really like the specific details the author provides about the interview set-up and the results observed among candidates at his company.
Now amazing as it is that there is anyone around who has not already heard of FizzBuzz as a job screen, I actually heard an even more amazing story in August 2012, which I shared here on Hacker News a month ago.
http://news.ycombinator.com/item?id=4603414 http://news.ycombinator.com/item?id=4603414
The story, which no one here seemed to believe the first time I related it, but which I fear is very likely true, is that the hiring supervisor who told me this story is the lead programmer for a company that uses Web back-end programming to serve up online, adaptive lessons. He has a position open for a new programmer as the company's business grows. He has advertised in the expected places like Stack Overflow and Joel on Software. During the year 2012, the company has received many resumes. Applicants who have computer science degrees, and especially the many applicants who have master's degrees and who otherwise look all right, are invited to do a phone screen in preparation for an in-person interview. During the phone screen, the applicants log into an application that allows them to type on their own PC to write code that can be seen by the hiring supervisor. The test the hiring supervisor uses is "Write a program using a for loop to print the integers from 1 to 100." Applicants are permitted to use any language they please, and the supervisor said he would even consider correct pseudocode a successful answer. He has not yet had ANY applicant, not even one, submit a flawless working program to do that. Not one. He was eager to interview an applicant who got the test MOSTLY correct, after seeing many applicants fail to do it at all.
I know the company that that hiring supervisor works for. I do not know how much its location or the fact that the software roles in that company are support roles rather than producing the company's main product may influence the desirability of the company. But somehow I think he is not lying when he says that a straightforward implementation, no tricks, of "Write a program using a for loop to print the integers from 1 to 100" is too hard for many CS graduates. That's why everyone hiring today should build a work-sample test into the hiring process.
http://news.ycombinator.com/item?id=4613543 http://news.ycombinator.com/item?id=4613543
It's not clear that higher education credentials guarantee ANY minimal level of coding skill, however elementary.