[lkxijlewlf]

> If you answer immediately, you will "ummmm uhhh" a lot. Learn to take a breather and buy time.

Not just in interviews (I don't interview that often), but just in general, I've trained myself to smile and say, "That is an interesting question!" (and subtle variants so I'm not a robot). It buys me a second or two if it's not that hard, and even more if it is a hard question because now they know I'm thinking about the answer and, well, that is okay.

Edit: And it often buys you points, so to speak, with the person asking.

[lowercased]

> And it often buys you points, so to speak, with the person asking.

I had a tech interview with some folks... 15 years ago. Smallish agency, and I was meeting with the owner and his #2.

I got asked some question - "how would you do X?" - I think it was something like "build a house". Not a tech task, just "how would you go about X". I went to a whiteboard and picked up the marker. Just before I started to draw something, I asked some questions. "Who's going to live there? How many people? Do they have any specific needs?" Stuff like that. Just a handful of questions, and I started answering/drawing based on their feedback. I asked a few more questions, got more answers, drew some more and explained things, then sat down.

I got a job offer the next day (turned it down - couldn't afford me). But I was told (both by the owner and later someone else I met who worked there) that I was the only person they'd interviewed who'd ever asked any clarifying questions before answering.

[scrozier]

Seems like a good technique, but boy, is that overused these days. Let's face it, not all questions are that interesting. Seems that a lot of people start with that, robotically.

[gknoy]

I can see how it might get old hearing that a lot, but when I look at the suite of tech screen questions we have here, several of them are basic under the hood (tree-building and traversal, string manipulation, etc). However, they all do have something about them that made them interesting to me when I did them to calibrate my interviews.

Things can be interesting without being hard or novel, and as Aphyr's entertaining "Hexing the interview" series shows, you can often find something interesting for purely personal reasons that are separate from the question itself. The interesting part might be that a normal data structure isn't enough, or that the difference between the naiive solution and the performant one is substantial. It might just be that we might be doing a task that we normally use a library for (string manipulations) in our normal work, so it's refreshing/challenging/interesting to look at things from a different level of abstraction.

[lkxijlewlf]

> Seems that a lot of people start with that, robotically.

Can't tell if you're trying to be an ass or not.

People can use whatever they want. The point, as many others point out, is to come up with something you'll use to not say, "ummm..." so much.