Training New Students in Undergraduate Research

This past week I was training two new research students. For new students in my lab, I set aside a day and a half to teach them some of the basics so they can get started delving into a research project. The learning curve can be steep because it is rare for a student to have had a class in computational chemistry before joining my group. I try to take students at all levels; one of my students took quantum chemistry last semester and will be learning some thermodynamics and kinetics this coming semester. The other student hasn’t reached those classes yet, but has a strong science background.

Over the years I have developed resources to help shepherd my students through the first stage. First they go through a Unix tutorial to familiarize themselves with the operating system and working in a command-line environment. Then they learn how to use the basic text editor “vi”. After that we cover the general setup of the hardware, how to log in and out of machines, check the load on a compute node, and other system-specific things. All this takes about half a day. As homework they then read one or two papers related to their assigned project, that also covers some of the basics of computational chemistry. I tell them that they won’t understand most of the paper the first few times reading it, and that’s okay. They will be re-reading the papers over the semester and comprehension will come with time.

The next day we go through the basic outline of how computational chemistry relates to things they’ve seen in class. We cover some basic terminology and students get a rough sense of their projects. I assign a project to each new student, because in the early stages a student typically does not have the background to outline a doable project. I do take into account their interests, prior knowledge and skill set. A student who has been in my group for a while may later shape their project according to their interests.

Then we start to learn the software. There are many steps to setting up a calculation for the type of work that I do, so we go through these step by step. While I have a group wiki that the students can refer to (and I write up “cheat sheets” for reference), I’ve found that being present and guiding a student through the first few runs is what works best. First the student sets things up with my instruction, line by line, on a “toy problem” (i.e. one for which the answer is known). Then we repeat the process for a slightly different calculation with different parameters, but still following the same overall methodology. This time, the student works it out (they take copious notes the first time in their lab notebooks) while I’m there to guide them when they get stuck or stumble over some of the steps. Finally we do this a third time where I leave the lab, and the student comes by my office if he or she runs into trouble midway. By the end of the day, the student shows me the results of the third calculation. I then give the student a little more reading related to the actual project and now the student is ready to delve into a project. I had forgotten how much talking I do during the day and a half and as a result I had a scratchy throat by the end of the second day. Thankfully it was a Friday and having avoided talking much in the last 24 hours I’m now recovered.

For the first half of the semester, my meetings with new research students usually take the form of my telling them what type of calculation I would like them to run, and their showing me their results and subsequent analysis. In the meantime I try to teach them a little theory along the way depending on their background and what classes they have taken. By the second half, hopefully they will start to have a bit more self-direction and figure out what to do next, i.e., I ask them what they think they should try next. Some students are quite independent and start to catch on, but others need hand-holding for a longer period of time. This is something I still find hard to predict even though I’ve had a fair number of students over the years. My research group typically has 3-4 undergraduates.

Being in a liberal arts college with no graduate program, means that there are no graduate students. Thus, the nature of undergraduate research is quite different (at least in the sciences) at a liberal arts college compared to a research university. For starters, the main goal of my running a research program is to give undergraduates the opportunity to learn how to do actual scientific research. They work on real problems (that no one knows the answer to), although I have usually carved out their projects into bite-sized pieces. My goal is not to become a famous scientist, nor do I have the time, energy or resources to compete with labs at research universities with armies of graduate students and postdoctoral researchers. To keep my lab running, I do have to write grants for extramural support and keep publishing in peer-review journals, albeit at a lower rate than expected at a research university. Thankfully, things have gone well so far, but funding gets tighter as the years go by, and the competition gets stiffer.

One positive outcome (and there are many) of undergraduate research is that students become more invested in their other science classes, at least that has been my experience. When starting on research, I think they get a glimpse of how much is unknown, and how important it is to keep learning. I try to make connections between their research projects and the classes they are taking. (There are always connections!) It’s a pleasure when students report that something clicked for them in class because of their undergraduate research project. Many of my students are co-authors on my peer-reviewed publications. This I’ve found is a strong motivating factor for them to work hard on their projects! Providing students the opportunity to present their work at national conferences is also a delight, and they usually find the experience mind-blowing. I’m particularly pleased when my undergraduates are mistaken as graduate students at poster conferences (we put them in the thick of it with graduate students, postdoctoral scholars, and full-fledged scientists).

Having undergraduates work in my lab also forces me to keep being involved in research. My time gets easily taken up by teaching (which I love) and administration (which is part and parcel of the job), and sometimes research gets short shrift. The funny thing is that if I wanted to be more highly productive research-wise, it’s actually better for me not to take on undergraduate students. I can do the work much faster than they can (at least the type of work that I do) and I would save time having to explain the ins and outs of the research project, methodology and the theory behind all of it. Not to mention, I have to check all their files carefully before I publish anything, sometimes checking and re-running calculations myself. So one might even say it’s double work on my part. But that’s not the point. The point is the education of the student and so I’m happy to have the students participate as co-creators of new knowledge!