This is the first I get to review PhDs applicants and would love to get some advice from the group. We all know well that it is a huge commitment and investment for a five years duration. I am so afraid to admit someone who doesn't fit me. Love to hear some words…
Date: 13 Dec 2014 02:06
Number of posts: 8
RSS: New posts
In no particular order:
1) Ask for advice from more experienced faculty.
2) Have applicants meet multiple faculty and students on the interview day. Listen carefully to their comments. It's easy to get blinded by someone's interest in your work.
3) Call the letter writers for your top choices. People are more honest over the phone and you can learn a lot from the tone someone uses.
4) Only take someone who you feel is exceptionally strong. It is better to not take a student if you're only options look bad.
I'm not sure there is any "magic formula" for picking a good graduate student. To illustrate - I was the only person/program to admit someone who has been not only my top student, but is a top student in my program (co-authored a couple of papers with me, including one in Psych. Bulletin, and just received a revise resubmit, minor revisions, on which they are first author, and is otherwise making the program look easy, in part because of their positive attitude and willingness to work hard). That said, think about some of the following:
1) is your vision for a lab one wherein team work is critical, or do you envision students working mostly on their own projects? If the former, keep that in mind in selecting a student - and pay particular attention to what other graduate students already in the program say
2) if you can't see yourself seeing and working with someone nearly every day for 5 (or more years), then I would avoid them. It can be hard to tell in a short interview, but you can usually tell the ones that you can't see working with, and then go from there. It is impossible to really tell from the paper applications, but look at what letter writers are both saying and NOT saying…..
3) try to find someone that is willing to work hard - easier said than done. That acknowledged, I'm willing to work with anyone to help them strengthen areas of relative weakness, but they have to "want it".
4) When first starting off, you might try to get someone who is interested in doing exactly the work you are doing or very similar types of work that you are doing. It will help you get your lab up and running, and help to develop a shared vision for the work your lab will do.
5) Try to interview as many as possible (within reason), but don't necessarily use all of your invitation spots (if there is a limit) on those with only the highest scores and recommendations from PIs in well-known labs. When first starting out, you and your work are not as well known, and those applicants are likely going to be admitted to other programs (I'm not suggesting you don't try for some of these applicants, but if you only invite these top applicants, you run the risk of not having an offer accepted). However, the "group" of applicants just below that top tier group can be quite good, and there may be some solid applicants that otherwise look "marginally acceptable" on paper. You often won't know until you meet them. Moreover, at the end of the day, in my program, we've had really top applicants on paper and during the interview come and end up leaving after a few years - every program has had this experience.
6) To the extent that you can, try to find evidence of solid writing, and good statistical thinking. I would take such a student with 70%-range GRE scores and a 3.6 GPA any day over a 90%-range GRE scores, 4.0 type of student without evidence of solid writing/statistical thinking skills any day……
Others will have different opinions on some of these things, but I hope those thoughts help you a little bit. It is a very important decision - a first graduate student or two in the "decent" to "great" range can really help you get going, but one falling on the other side of the spectrum can slow you down substantially.
If it were me, I'd only consider applicants who have demonstrated background with conducting research - as a research assistant or, better, doing undergraduate research work such as an honor's thesis. Then, for those applicants I'd contact their recommenders, one of whom should be the research supervisor, and ask questions. Going directly to the source avoids possible cookie-cutter recommendation letters that aren't all that accurate.
Yes, I would not take anyone coming straight from undergraduate. It's just a sample size problem: Someone who has only worked with maybe 1-2 people for a few hours a week as an undergraduate simply doesn't have much of a track record. They could excel partly through random chance. It's like drafting baseball players: You may be able to get some gems by drafting high schoolers, but drafting out of college is a much more sure thing.
Personally, I'm preferential to folks from a strong SLAC. They tend to be under-valued, in part because so many people in our field are foreigners and are unfamiliar with SLACs. But my general experience is they tend on average to be more dedicated than, say, Ivy League graduates. Certainly, their coursework is more rigorous, to the extent that matters.
I sure am glad my graduate advisor had lower standards than all of you, or I'd never have made it anywhere!
I think it is very important to be able to 'cut the string' and get someone out of the lab/program if you believe they are not going to work out.