A question for the PI’s among you: how selective are you in choosing prospective graduate students?
To kick the conversation off, I’ll share my own answer. I’ve always been pretty selective, meaning that I only take on students whom I’m confident will be very good fits and who will do very well. That’s for several reasons:
- I don’t want (and on my current grant, can’t afford) more than 3-4 graduate students, which is probably on the small side for an ecology lab group at a research university. I don’t need lots of bodies to pursue my research program, and I want to be able to give each student as much one-on-one time as needed.
- When I first started out at Calgary I wanted “my” research program to consist largely of “graduate students pursuing their own projects”. That was for several reasons. I believe that one of the most important things you can learn in graduate school in ecology is how to think for yourself and develop your own project.* I thought it would be a good way to start up new lines of research that I might want to continue pursuing in future. I’m in Canada, where the funding system works in such a way that my students don’t have to work on specific projects or lines of research that I’ve designed. And I knew I’d have trouble attracting graduate students to work in my own system, since it’s lab-based and few ecology students want to work indoors.** If you’re going to give graduate students a lot of independence, and you’re not going to take on many of them, then you need the ones you do take on to be quite good students who are very good fits for your lab. Otherwise, your lab won’t be very productive and it will be difficult to continue to get funding, even in Canada.
- Even if you make a point of getting to know prospective students really well (which I do–I don’t ordinarily take on a student who hasn’t come out to Calgary for a one-on-one visit), there’s always some risk that things won’t work out well, or at all. So that you as a supervisor will have to put in a lot of extra time and effort trying, and possibly failing, to make things work. And a lot of that time and effort will be no fun. As a supervisor, I dread that possibility. So to minimize the risk of a student not working out, I’ve tried to only take on students whom I’m quite confident will work out.
- I think I’m a pretty good recruiter. When I’m looking for prospective students, I advertise pretty widely, through various channels (EcoLog and Evoldir posts, emailing colleagues, posting on the blog, posting on my lab homepage, etc.). I have to be a good recruiter because for various reasons I don’t have armies of prospective students beating a path to my door.
Being quite selective has worked out very well for me so far. I’ve been fortunate enough to attract several really great graduate students. But lately I’m wondering if I need to be less selective. Due to a combination of graduations and transfers, I’m now down to one graduate student. And in the last two years I’ve had several promising prospective students decide against joining my lab.*** So there’s some non-zero chance that a year from now, my current student will have finished, leaving me with no graduate students until fall of 2015 at the earliest. Which wouldn’t be a disaster, but would be quite sub-optimal.
Aside from concerns about my lab size, I’m also wondering if I need to be less selective in the sense of being more willing to take on students to whom I will then hand a partially- or completely-designed project. Thanks to the work of my first cohort of students, I now have a couple of new lines of research I’d like to continue pursuing. So while I’m still open to taking on students who want to pursue their own ideas, it’s now more important to me to also take on some students who will pursue research projects I’ve already decided to pursue. The students pursuing those projects would still have some independence, but within narrower bounds.
Perhaps further down the road I’ll be able to be more selective again if my circumstances change. In particular, my lab has yet to publish anything on the new lines of research that my students have developed (hopefully soon!) Once those papers come out, I’ll be in a much better position to say to prospective students “Here are the kinds of research you can pursue in my lab.” At the moment, any student who wants to do anything other than protist microcosm work or mathematical modeling would be taking a bit of a leap of faith in choosing to work with me, which most prospective students understandably are reluctant to do. I was fortunate early on to have more than one grad student who chose to take such leaps of faith.
But enough about me. How selective are you about prospective graduate students? Why? (I’m betting that the vagaries of different funding systems will matter a lot here) Have you become more or less selective over the years, and if so, why?
*I think that’s true whether or not the student plans to go on in academia.
**This is why I just laugh on the (fortunately rare) occasions when I hear a field ecologist complain or worry about how we don’t have enough field ecology students because lots of students are getting sucked into microcosm work. And it’s why my microcosm research at Calgary has mostly been done by undergraduate summer assistants and the occasional independent study or honors student.
***And I completely respect their decisions and wish them all the best!
Dynamic Ecology 
Greetings Jeremy,
Not an ecology lab here (or at least not intentionally – we’re 80% computational biophysics, but spend a lot of time puttering around in species/individual/genome/phenome/environment interactions in polymicrobial colonization of various niches, so we cross paths with microbial ecology occasionally – anyway, I digress, but divergent field aside, I still find a lot of interesting thinking points in your posts), but I’ve been pondering graduate student recruitment for my group a lot recently, and your post has just caused an epiphany:
My recruiting strategy has always been almost the polar opposite of yours. Not in being non-selective, but in that I’ve always prioritized students that I /wasn’t/ sure were going to be guaranteed successes. This is partly because I feel a lot of civic duty to my graduate program, and someone needs to pick up the stragglers that enter the program, but can’t find a lab that’s a perfect fit. More, however, it’s because I’ve always wanted to find the hidden gems – the quirky brilliants – the kids who, regardless of whether they were going to tank or succeed, were going to do it in a stellar fashion.
By and large, this has historically worked for me, but like you, in recent years I’ve been succeeding in recruiting fewer and fewer of these into the lab, and I’ve been wondering if I needed to re-think my approach.
The epiphany I’ve had, is that perhaps it’s not that your approach is sub-optimal, or my approach is sub-optimal, but that maybe “a” approach is sub-optimal. I have to admit that, despite trying to avoid complacency, repeatedly trying to attract the same type of student has become somewhat formulaic. Despite the fact that every student ought to be an independent experiment, perhaps it’s that consistency that’s hurting our recruitment.
Re-thinking things in the light of your post, and an inadequate dose of coffee as yet this morning, I am wondering whether re-focusing my priorities regularly might force me to keep each recruitment more personal, and improve the chances of success. More coffee will be necessary for pondering this…
“someone needs to pick up the stragglers that enter the program, but can’t find a lab that’s a perfect fit.”
That’s a big difference between your program and most (not all) graduate programs in ecology and evolution. Most EEB programs won’t admit you unless you already have a supervisor lined up. So I’m not selecting from among students who’ve already been admitted to our program.
“perhaps it’s not that your approach is sub-optimal, or my approach is sub-optimal, but that maybe “a” approach is sub-optimal”
Yup, that’s exactly what I’m thinking too.
As each year goes by, I become more selective.
I’m at an institution that only takes on MS (or for non-US people, that’s MSc) students. We get a mix of students who are looking towards a PhD program, those wanting a health sciences career but with a lackluster undergrad background in their way, some going into industry and some high school teachers.
There is one person in our department who has been very generous with his time and has taken on the students who otherwise would have washed out of the program and nobody else would take them on. It’s been good for those students, but has really dragged both him and his lab down, and prevented him from working with students that would benefit more from his high-quality efforts.
I have only been moderately selective over the years, and would take on academically talented and hard working students regardless of their career goals. Now, with a growing sample size, I’m discovering (or at least coming to the opinion) that career goals are highly predictive of thesis success. No matter how smart and hard-working, a to do a great thesis in ecology, a person really needs a person who is excited about ecology. It’s not like students are beating down the door to do a MS thesis in my lab, so being more selective will mean having fewer to no graduate students. But that might be a better choice for everybody.
Having literally, this morning, just offered a student a PhD position, as well as another one earlier this week, this post and comments are pretty well timed!
In both cases, we advertised quite widely, had to turn down interested, otherwise well-qualified applicants (due to the University scholarship being restricted to EU nationals/residents) and were very selective. Having a high calibre of candidates made that part easier. The positions are linked – (1) a lab based project with (2) theory being developed around that. I’m acting as a “bridge” co-supervisor between the biology and maths departments. This had an effect on the type of applicants we had – biologists, mathematicians, physicists & engineers, an aspect that was particularly exciting for me! It also meant that while the project is generally defined, there’ll be space for the students to go off in their own direction (e.g., extending things to the field).
In terms of numbers, we got more than twice as many eligible applicants for the “theoretical” position (12) as the “lab” based position (5), which surprised me – although it’s admittedly a small sample size. This compares with ~18 applicants for a field position we advertised last year.
This year, I tried to be quantitative in the interview process, awarding marks out of 5 for each response. I think this helped me compared to last year, where I just made brief notes on the candidates’ responses, which I didn’t find all that useful when reviewing after speaking to them all.
Anyway, there’ll be a bit of a lag before I can report the success/failure of these approaches. My only existing (graduated) data-point did very well, based on a gut-feeling appointment by the other co-supervisor. He tended to recruit that way, and was pretty successful over a long period!