I get such a noisy signal as to whether a given technique or answer went over well (Okay, I didn't get that job. Which of the 400 answers I gave was wrong?) that I'm hesitant to give broad spectrum advice. You can get a lot better answers elsewhere (see sidebar for examples). But all the good, tested advice comes from HR people, and there are some things they just don't get. So here are answers to questions from technical interviewers
The Question: I bought a car from some guy. He's given me two days to return it. I give it to you to test. What do you do to make sure I don't get ripped off?
The Answer: This obviously a multi part answer. But the bit that set me apart (and I know this because he specifically told me) was that one of the steps I listed was to look for documentation on-line. Previously I'd worried that saying that made me look lazy, but apparently it looks efficient, which is just like lazy except employers like it a lot more.
The Question: can you develop GUIs?
The Answer: Yes, but they will be ugly.
Obviously, that doesn't apply if you're being hired as a graphic designer. But if you're a tester or dev, and you know they're specifically asking whether you can design interfaces for in-house tools, this is the only answer that will get you respect.
And now, a technique that I've gotten less feedback on but really seems like it ought to work:
When you're listing test cases, list the circumstances under which they would and would not be valuable. For example, in the car case above, I mentioned that because I was testing a single car for personal use, I would not test to destruction, but if I was the QA Engineer at Ford, I would select some number of cars to push to the absolute limit. This helps you two ways. One, it shows you think about the problem and the best way to approach it, and don't use a one size fits all attack. Two, it fills up time while you think of more appropriate test cases. If the choice is between dead silence while you think (and if you're like me, the harder you think the more bored you look) and an inappropriate test case, labeled as such, go with inappropriate.*
*Note to the socially challenged: this does not generalize. Silence is almost always better than inappropriateness.
Thursday, October 25, 2007
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment