Advice to Prospective Grad Students

Professor Singer's Advice to Prospective Graduate Students:

Here is some advice for folk in thinking about grad school, intermixed with comments about grad school at U Texas and my own experience as a supervisor. Don't be in too much hurry, postpone your own decisions as long as possible so as to have the maximum of information at hand, this is a complex decision! (see below)

1) Choosing a school: You can choose to go to a school where there is a specific person with whom you wish to work, OR you can choose to go to a school that you judge to be excellent in some more general sense. The first has its disadvantages in that the supervisor may disappoint you, and there is nowhere to go without switching schools. Best to have both, of course, a person with whom you'd like to work in an excellent institution, there's not necessarily a conflict. Different schools may offer you very different financial deals. You may be faced with a choice between a miserable deal at a place where you'd like to be, and a much better deal at an institution with less appeal to you. This is a nasty decision to have to make. If you have a high GPA you can sidestep this problem by matching it with a high GRE, then getting a NSF predoctoral fellowship. These are quite lucrative and you can take them with you if you switch schools. You can get them either before entering grad school or during your first year there. If you have one already, your chances of entry to the grad school of your choice are improved. These fellowships are awarded mostly on NUMBERS, so if your GPA is low, that doesn't mean you won't be good at research, but it does mean that you won't get one of these fellowships. If your GPA is very high, it's well worthwhile to put effort into the GRE scores (verbal, quantitative and specialist subject) for this reason. Don’t assume that an institution that has an excellent reputation in general will have appropriate expertise for your own interests, you need to check up on that in detail!

2) Choosing a supervisor.

2A.. Independence

First, you should think about whether you prefer to work independently or in collaboration with an individual or group. This can be a tough one. There have been some stunning examples of totally independent dissertations here at UT in recent years. Successful independence brings genuine self-confidence and makes you look a good prospect for an academic career, if that's what you want. It doesn't matter as much if you want to go into Fish & Wildlife, National Parks, or NGO work. On the other hand, independence is more risky scientifically and harder emotionally, 'cos, with the best of goodwill, you will need to depend on folk in other institutions for genuine enthusiasm about your work. Your supervisor won’t have the time to become expert in what you are doing, if it bears little relation to their own work. The independent student has an increased risk of planning and carrying out experiments that don’t mean what they seem to mean, because of lack of experience and intuition about the study system. If you have a supervisor who works, say, on box turtles, and you have a conceptual question that you’ve thought about, you should be able to say “Hey, supervisor, suppose I were to try this experiment with your turtles, will it work?” And you should get an accurate answer, the turtle biologist should be able to tell you whether, and how easily, those box turtles would answer your question.

What are the pluses and minuses of collaboration? Dependence on a well-funded supervisor can be a fine situation because your supervisor can pay you Research Assistantships and take you out of teaching, which, again, gets you moving faster and enables you to do field work at any time of year. Here at U Texas, you can very likely get money for the research from the Dept if your supervisor has no money (or no money that can legitimately be used for your project). These Dept funds might pay for your field expenses but they won’t pay to take you out of teaching!

The dependence on a supervisor that is inherent in collaboration also carries risks, at least two serious ones.

The first risk is that your supervisor may be misguided. One of my best former students, Chris Thomas, came here to work with Larry Gilbert, but then he read one of my papers (1983) and came to see me. He said (and I quote verbatim):

"Dr Singer, I read your paper in Evolution, and I thought it would be very interesting IF TRUE. However, I did my MSc on butterflies and I'd be very surprised if you could really do that sort of work with them. But IF you could, I'd be very interested to follow it up for my PhD. So, what I propose to you is this. I'll go to California with you on your field trip this summer, provided that I can spend the first field season double-checking your published work."

Well, this degree of explicit honesty may not always be the wisest policy, but the sentiment he expressed is indeed a wise caution. I've seen one or two PhD's fail terribly because the student had too much trust in the Professor's prior work. And professors can be wrong just because of bad luck, without being dishonest.

The second difficulty with collaboration is that, again, if you wish to look for academic employment with your PhD, you need to avoid looking as tho' you have spent 4-6 years effectively working as your supervisor's underpaid technician. But, on the plus side, if the supervisor is competent, you get stuff achieved faster, especially at first.

There's no clear answer to these questions, the best strategy for you depends a lot on your own personality. But I mention it now because it may influence you choice of institution. Not all institutions or supervisors welcome independent students. If independence is your aim, you need to find out if this aim is respected. Here, it's no problem either way. About half my students have been independent. However, it seems to me that, again just among my own students, those who have done best academically are those who have started out initially in collaboration, have used this to learn how to work with my study system (Melitaeine butterflies) and have then subverted the system to their own ends, so that they ended up obviously independent because they had tackled conceptual questions clearly separate from those in which I had historically been interested. Over time this process has often ended up changing my own interests. I am now interested in patch dynamics and metapopulation structure because of my collaborations with two students, Chris Thomas and Davy Boughton, and this interest has brought me into collaboration with Ilkka Hanski's group in Finland and Isabelle Olivieri's group in France. (You'll see papers by Boughton & Thomas on this topic in my cv).

2B Choosing a supervisor: What kind of person???

I strongly recommend that you read papers before choosing a supervisor, and ask yourself if the approach and the way it is described appeals to you esthetically as well as being scientifically correct. Also the papers will tell you if the prof's students are readily able to publish papers that do not bear the prof's name. Single-author papers are not essential, but can be very useful evidence of academic independence (that is, if you're in non-molecular ecology behavior or evolution, rather than in molecular bio, where the supervisor's name almost always goes on the end, for getting the money and running the lab). When you get down to choosing between a small number of potential supervisors, it's not a bad idea to talk with some of their past as well as current students.

Way back in 1978, one of my grad students stood at my door and said, crossly:

Mike, look at yourself! Academically, you're going down the tubes fast! How long is it since you were in the LIBRARY, huh?"

Well, I admit I was a bit taken aback by this, but it materialized that she was worrying that I wasn't getting famous fast enough to write influential letters of rec for her when she graduated, and she was an academically ambitious person. In retrospect she was right to worry about this, it matters more than it should. But if you choose an excessively famous supervisor you often can't get to talk to the person. So, the question from your perspective is how to optimize the tradeoff between faculty fame and faculty accessibility, both of which are desirable traits. I think that the ideal situation is to find a young supervisor who's with-it in terms of what's going on in the field of research and on the way up in fame, but not YET so famous as to be inaccessible. This might be quite a trick, but I've observed that the very best young faculty often go without students for years because the grapevine is too slow, students are being advised to go with faculty who are known to the folk doing the advising, who are out-of-date.

So, talk to prospective supervisors, size them up and read their work. Don't be too tempted to work with those who are more willing to put themselves out to talk with you, the folk who can most easily find the time to do that are those who are not doing much research (or who are terribly efficient!) Be completely selfish, do what you feel is likely to be best for yourself, both personally (ie, in terms of lifestyle) and professionally. Some folk may have been NICER to you than others, but you should feel no guilt at turning them down, this decision is enormously more important to your future than to theirs! You should evaluate their personalities only in as much as they would affect your working environment and your enjoyment of your work. If relationships with supervisors are an issue, it might be useful to contact the most recent past students of potential supervisors. Those who have done well will be easily found, those who are not employed in biology will have effectively disappeared. I expect that all of us will have both categories! I certainly do.

Choosing a project;