Two key litmus tests when devising questions: Is this relevant? And is this diagnostic? To explain what I mean - in my line of web development, here are some actual technical questions I've been asked:
- BAD: "What is the syntax for converting a DateTime object to mm/dd/yy format in C#"?
- OK: "How does Ajax work"?
- GREAT: Compare and contrast the popular data access methods out there
The "OK" question is good on the one hand, because the interviewer is trying to dig to see exactly how knowledgeable the candidate is for behind-the-scenes components in the web request/response protocol (have I lost you yet?), but on the other hand, who cares. We've got frameworks that handle the nuts & bolts now. I don't have to know anything about the XmlHttpRequest() object. In fact, if I see it in your code, you're writing bad code, because you should be using a framework. It's diagnostic, but not.relevant.
The "Great" question requires the candidate to really know his/her stuff (e.g in .NET world: LINQ, DALs, etc.) and even better, to diagnose the strengths & weaknesses of each, ultimately form an opinion, and then be able to articulate that opinion concisely in an interview. Hard. But both relevant and diagnostic.
Stay away from multiple choice questions and trick questions. And if a lunch interview, don't chew with your mouth open. Happy hiring.
No comments:
Post a Comment