Interviewing a prospective bioinformatics graduate student? Here are a few tips.
- Keep your lab's research web page up to date. The key to getting a good student comes before the interview begins.
- Know a bit about the prospective student. You can expect the prospective student to know a bit about your work (so long as you fulfill tip 1). You should take an interest and know something about the prospective student you're interviewing. This sets a very positive tone to the interview and makes the student feel welcome and comfortable. I had excellent interviews with professors who took the time to read the dossier the application committee made of me. (If your application committee does not prepare such dossiers, make certain it does next year.) As an addendum, if you know a colleague at a prospective student's previous place of education, feel free to ask if the student knows the colleague, but do not expect them to say yes.
- Talk about your research first. Sure, the student who is really motivated will have already read your research web page. (See 1.) Regardless of preparedness, however, most prospective students will still feel nervous going into your office and sitting with you for the first few minutes. Ease the tension and break the ice by talking about your own work before having them talk about theirs. If you allow that buffer time of several minutes for the prospective student to calm down, the interview will go much better for both of you. From personal experience, my best interviews started this way. I felt most welcome by professors who could also articulate exactly what areas they hoped a grad student could step in and help. If you allow that buffer time of several minutes for the prospective student to calm down, the interview will go much better for both of you.
- Don't ask the prospective student to code a particular function in <insert favorite programming language>. A prospective student can take offense at such a request. It shows a distrust in the skills they claim to have and immediately puts the prospective student on the defensive. Also consider the pointlessness in the exercise, which places an overemphasis on coding skills. In my experience thus far, programming is but a small part of a bioinformatics Ph.D. It is a skill that can be improved with remedial courses and Q&A among fellow grad students. Instead, focus on asking questions that probe what really matters: the student's interest in learning and self-motivation. If the student has a particular interest in programming, then sure, let the conversation wander in that direction. (For instance, Prof. Paul Magwene and I had a kindred-spirit talk about our love of Python.) In general, though, take a student's word that they program in the languages they claim, and if they don't program at all let them know that they'll need to take remedial courses to learn and check that their reaction is positive to this.
- Ask the student about his or her research. The only way you'll really get a sense for the student's propensity for research is asking about his or her previous experiences in it. The more specific your questions about the student's research, the better. (See 2.) An example of a terrible question would be, "What was the greatest challenge in your research project?" An example of an good question would be, "It sounds like you worked with several different phylogenetic models. Were the resulting trees similar, though?" The former question is too open-ended (much like my most hated question of "Where do you see yourself in five years?" [Note: Best answer, "In the mirror, of course."]) The latter is specific but still leaves room for the student to explain his or her findings.
I'd like to thank the Biogang on FriendFeed for inspiring this post.