Monday, July 19, 2010

What I Look for in a Graduate Student

Here at Prodigal U, it is the season for recruiting students. In my department, we admit students without an adviser. Students arrive in August, and meet with different professors to choose an adviser (and select appropriate classes for September). This is a relatively common method in my field, although at other universities (like PhD U), students pick advisers after a semester (and TA for support). Some (many?) students arrive already knowing who they want to work with, and some have already arranged it with their future adviser, which is perfectly fine. As a new prof, I ask for meetings with all the students interested in my research area, since I want to see as many people as possible to find good fits for my group. After last year, I have a much better idea of what I am looking for. Last year, I kind of did this a little "seat of the pants", and lucked out. This year, I have been thinking a bit more about recruiting ahead of time. So what am I looking for?

1. Previous research experience. It is really hard to tell if a student who has a high GPA will be good at research. If they already have some experience, I can get a reference from their previous mentor(s). I can also ask them to talk about their research to get more information about how they think about research and science. They also have some experience with the difficulties of research, and have more of an idea about what life in the lab will be like. I need fast starters, since I am just starting out, so this is a big one.

2. Good to great GPA. Grad students need to get a B in their classes to get credit towards a degree here. I want students who know how to study, learn quickly, and are motivated enough to do a good job on something required for their degree (even if they hate classes). I also don't want my students to be so absorbed in passing their classes that they don't get going in the lab. I am totally fine with students who started out poorly, but did well as juniors and seniors. The opposite trend I would find disqualifying.

3. Self-motivation. I am still working out how to select for this. I am a very hands off manager. I don't like to micromanage, and I don't want to have to enforce working hours/face time. I want students who like this kind of workstyle and can work efficiently in it.

4. Works well with others. We do a lot of collaborative work. Students looking for the lone wolf at the bench experience won't get it in my group, and I don't want headaches from territorial drama if I can avoid it. I talk about working with others up front. I also like to ask students about group projects they may have worked on--what they did, what the goal was, if the goal was reached, and where the problems were.

5. Excitement about working on a research project. I don't need my students to be super-peppy or anything, but if they look and sound bored when considering the possibilities in my lab, I will assume they are bored, and my group is not a good fit.

Last year I was lucky, and had 4 people interested in joining my group, of whom I took 2. I have been in "addition by subtraction" situations at work before, so I definitely know I would rather have no one than a poor fit for the group, especially now that the data is flowing nicely. I would like to add 1 or 2 more students this year, and then I will be at the limit of what I can afford without kicking the external support up a notch or 2. Any suggestions for other things to think about when recruiting?


Odyssey said...

Excellent list of requirements. Number 1 really is the number 1 requirement. We have it pretty much as an absolute requirement for students to get into our program. That might seem harsh to those coming from small institutions with limited research opportunities, but the cost of having a grad student drop out because they discover they don't like doing research is just too high.

Gerty-Z said...

This is an interesting post. I'm going to be getting my first students this year and I have been wondering how to screen through them. In our program, all students do rotations. I don't think (and it was not true for me in gradU) that a large fraction know where they are going to end up before they arrive. And those that do "know" are often surprised and end up somewhere else. Do you do anything to recruit students into your lab, after they are already in the program?

Anonymous said...

I think to some extent requiring students to have research experience also filters self-motivation. If research experience is not part of their course, they have to be motivated by science to seek out and experience research first hand, giving up vacations or free time to get some experience and they apply with the knowledge that they enjoy the trials and tribulations of research. I suspect it also filters the excitement about research too; perhaps not project-specific excitement, but excitement about performing research and doing real science.

From my point of view as a soon-to-be grad student, something I very much value in those around me is academic honesty - being very upfront about what you do and don't know. I find way too many people bluff their way through things not asking what they perceive as "stupid" questions. I think this is a really important quality in potential PhD students, although I'm not really too sure how to test for it!

prodigal academic said...

Odyssey, that is my thought on previous experience as well (although we don't formally require it, I think a student would be hard pressed to find an adviser without it).

GZ, here is what I did when I arrived last year to help recruit: I set up a website describing my research interests and plans along with my CV and pub list. I signed up to speak with all of the incoming students who were vaguely interested in anything close to my area. I set up a quick info sheet describing how exciting my research is and what students would learn in my lab. I posted this sheet outside my door, and gave it to every student I spoke with. I didn't go through the application files in detail until people expressed more of an interest in my group, but this year I am looking at the files first.

TT, that is an excellent point about academic honesty and not asking for help. I also don't know how to select for it!

GMP said...

Great post, Prodigal!

I wish my department did rotations like what Gerty-Z mentioned.

Ideally, I like to recruit from the courses I teach. I usually teach a couple of core grad courses and a couple of undergrad core. After the semester, there are usually one or two students whom I tap to join the group, or, in the case of undergrads, work with the group over the summer. I've found that after they have taken a course with me, I feel much more confident about taking someone on than I do sight unseen; and they know if they like me too. Although, since we have a lot of international students and not too many TA-ships, we have to bring most students into the PhD program on RA's, essentially sight unseen. I look for good grades in core areas, recommendations (if from people I know) and statement of purpose, but it's really hard to decide based on written applications without even brief 1-on-1 meetings. Prodigal, your list of requirements is a very good one, but of course no imaginable list of criteria is 100% foolproof. There is a fair bit of luck in finding a good fit student... Sometimes the ones that look so-so on paper end up being spectacular once they find their calling. Unfortunately, vice versa holds as well.

I also think it's fair to take a student on a trial basis, say 6 months, before making a permanent commitment.

Anonymous said...

I understand why you say what you do about wanting new students with research experience. I didn't have any when I applied to graduate programs and found out the hard way how much importance is placed on that, almost above all else it seems. I actually find that sad. I worked two jobs through my undergraduate degree because it was more important that I graduate with a minimum of debt than it was that I have a summer of research experience. Yes, you can be paid to do research, but I couldn't walk away from one of my jobs for two-three months and expect it to be there when I was done.

Lucky for me I was accepted to a highly ranked (at least in my field) graduate program at PhD U that allowed us one year to do rotations as TAs (the best way, IMHO). I have just finished my 4th year and should graduate in 2011.

To me, you also need to look for someone who's willing to think about what they're doing. While good hands will produce lots of data, that a PhD does not make. I think in talking to each student you can tease that out to a reasonable degree.

Perhaps the most important thing, though, will be how they fit personality-wise with the students you already have. It only takes one tech/student to really affect the atmosphere in the lab - and it can be a big effect, either positive or negative. From the student perspective, that is worth the extra time and really appreciated.

Your students may actually be able to find out additional information to help make your decision easier. If you bring the newbie around to the lab for a bit - leaving the new student there to talk to your current students - both will learn a lot. Also, the newbie will be more relaxed in the company of other students than with the faculty.

~Pharm. Sci. Candidate

Dr.Girlfriend said...

I totally agree with the above about a good fit. Someone who does not get along with, or is a burden to the group, can cause an overall dip in productivity and moral. I think it is essential to at least get input from the postdoc or technician or senior grad student who will be supervising and/or mentoring the newbie in the beginning.

The safest bet is always someone who has held down a full-time job for a year or more (either before of after undergrad). It says a person can work full time, time manage, and function as an adult - some people straight out of college can be very immature for their age, especially if they have had the luxury of just focusing on study and unpaid research. Grad school is a full-time paid job, and you want someone who appreciates that.

If a person has previous experience working in science (academia or industry) - all the better. Technical skills are always an advantage. An tech in a university setting will also have a good idea of what is expected of a grad student, and depending on the PI may have experience working on independent projects.

I am not saying that recent graduate who has only ever worked part-time or summers at the bench cannot workout - but you will always be taking a risk. Just one year as a tech can make all the difference in terms of ability and commitment because grad school is nothing like being an undergraduate.

prodigal academic said...

GMP, I also wish we did rotations. Unfortunately, it would be frowned upon in the department to accept someone to the group on a trial basis. Fortunately, I at least get to meet potential group members in person before deciding.

Anon, I hear what you are saying about research experience. Unfortunately, as a TT prof, I can't take the risk that someone will start out, hate research, and quit. At PhD U, I'd say that about half of the students without prior research experience quit in a year. At Prodigal U, many students in my department do research for credit during the semester, so working students have a chance to get some experience. I wish this were more common.

Dr. G, I would totally recruit students with industry or technician experience heavily. Most of our pool are either straight out of undergrad, or come in with an MS.

Physician Scientist said...

I had great luck picking female grad students. All were reasonably mature at 22-24 years old. The men were a different story as the maturity wasn't there. I finally decided to screen based on whether they were "gamers" or not. Someone who spends 12 straight hours on a weekend day with their online "friends" might not be right for my lab.