As a converse to this question, as a professor, I find it difficult to conduct a useful interview. What types of questions should I ask which would give me a good idea as to how (1) productive and (2) self-sustaining of a student the interviewee may be? If different questions are required for each parameter, please mention that in your answer.
Ask them about their expectations out of their coming 3-5 years as a grad student!
I find that a significant portion of the frustration that I (and those others around me) have experienced is due to severe disparity between expectations and reality. Note that this is as much for the benefit of the prospective student as it is for you as the PI.
Specifically things to consider regarding expectations:
- The amount and style of supervision the student expects. My expectation was that I would "do science" and not "paperwork", in reality my PI knew less about the paperwork then I did, so I typically ended up needing to do a bunch of paperwork regarding employment (like salary raises, progress reports etc) and always late too...
- The abundance and lack of relevant competencies in the lab, for that proposed project. I was confident that the lack of computational competency at our lab wasn't gonna be a problem for me. I couldn't be more wrong..
- The workplace interaction with colleagues. I was under the impression that we'd be a team of intellectuals, tackling problems together, from our own angles. Again I was dead-wrong... Here, everyone's buried under their own pile of... Similarly, I can imagine that if the student expects to be a lone-wolf, forced team-work might be frustrating in the long run.
- How to handle getting stuck/frustrated. This is quite self-explanatory I guess..