Research Student Job Interviews 101

I have interviewed dozens of research students over the years for student internships in R&D. Before that I interviewed several times for internships and student jobs. Here are some take aways from my experience.

1. Answers have to be straight and to-the-point. Researchers are fact-driven people who dislike spin, buzzwords, and anything else that is more fluff and less meat.
2. Giving short, concise answers leads to being asked more questions (since interview durations are fixed). This means getting more chances to impress the interviewer. 
3. You interview with a person or a group, not a company! Make sure you have something intelligent to say about the specific group when asked about your interest in the position. The whole world knows Microsoft is a great software company with $XYZ billion revenues and ABC thousand highly skilled employees. But can you provide some insight about the research publications coming out of the image processing group at Microsoft Research where you are interviewing? 
4. Know your CV inside out. This is very important because interviewers can pick out any bit and ask you about it. Anything on the CV is fair-game.
5. This also leads to the corollary of not exaggerating or lying on the CV. Research scientists or professors are highly trained people who can easily discern how much someone actually knows. A student with a glowing CV and ordinary interview performance will be judged much more harshly than a student with a ordinary CV and an ordinary interview performance. Don't dash lofty expectations of interviewers. Do this by not creating lofty expectations in the first place. 
6. I-dont-know is the best answer when you dont know the answer to a question instead of trying to dodge the question. But its also useful to add one sentence after saying I-dont-know to indicate any independent thinking on your part about how the question may be answered. This helps drive the conversation forward (see the next point).
7. Difficult questions are asked so that interviewees can estimate your independent thinking capability. This is probably the single most important capability for any researcher. Try to work with the interviewer to arrive at an answer through common-sense reasoning if he leads you on by dropping clues.
8. Refrain from political or opiniated statements (e.g. I hate Windows but love Linux). I once told an interviewer that I disliked C# because it was not open source; at the end of the interview he told me that his team used C# via an open-source C# version (Mono)! Check-mate.