By Audrey St. John, Associate Professor of Computer Science at Mount Holyoke College
When I first started teaching, I was mystified (and, frankly, at times panicked) at the thought of having undergraduates work with me on research. I realized this was part of the job, part of my institution’s mission, but I just couldn’t figure out how it would be effective. Sure, these students were bright, eager and motivated to learn, but how much could they contribute with such limited time? A typical research experience might be 8-10 weeks during the summer (full time) or 10 hours a week during a semester; best case, I might find a student who would work with me for a couple years in this way. I had just finished six years in grad school and still felt like I knew nothing. On top of that, my research is at the intersection of computer science and math with applications in the domains of engineering and biology – would I be able to find students with experience in even two of these fields? As it turns out, I would soon discover how powerful research with undergraduates can be, and I’d like to share some of the lessons I’ve learned over the years.
I remember thinking I should come up with a list of very specific problems, solvable with limited time and background, before trying to find students. Looking back, I think I was trying to mirror the familiar classroom experience, where a careful syllabus provides clear expectations to students with specified prerequisites. It turns out that specific problems, while useful in giving students an idea of the research area, almost never provide the direction we end up moving toward. I suppose I should have seen that coming, as research never does go as planned. It can’t be clean and predictable just because undergraduates are involved. I’ve learned to embrace the prospect of the unknown, instead looking for students with more broadly defined interests, such as computational biology or robotics.
Once I’ve found students, the most successful approach comes from guiding them along paths that suit their own passions and interests. Many times, they don’t really know these in advance, so I view the first part of my time with them as a chance to let them play with different types of problems. This may mean coming up with examples that fit a given set of combinatorial properties, reading and presenting a research paper on an algorithm we hope to generalize, or building modules of Mathematica code to explore properties of certain matrices. One summer, my students built little robots with microcontrollers and old VCR boxes; that activity resulted in one student determined to continue working on hardware and another determined to work only on software (a surprise to her). This “discovery period” can be truly transformative for some, and the reward of knowing that I helped a student find out a little more about herself is one of the main reasons I became a faculty member.
After identifying her interests, the student begins to get a glimpse of how research feels by facing the energizing and terrifying prospect of defining her own problems and pathway. I have the students pitch their own projects and timelines (which are always too ambitious) and work with them to create several milestones along the way. I usually let them start off on their overly optimistic timeline, but know they generally won’t make it past the first milestone. The students track their own progress by maintaining a website with blog updates on their work. This serves two purposes: (1) it helps me understand what they have done, what they understand and where they are stuck, and (2) it gives them something to reflect on at the end of the experience. Throughout this time, I am very conscious of each student’s confidence level. For some, the unfamiliarity of not having lectures, assignments and a textbook can cause them to doubt their own ability. Explicitly telling them that research is coupled with a feeling of the unknown and relating imposter syndrome stories of my own and of other researchers often gets them back on track. This is one of the things I enjoy the most, mentoring students who are excited to work on problems related to my research and helping them find the confidence to jump-start their own research careers. It is an amazing feeling when they tell me years later that it was that seed of a research experience which grow into their passion, whether it is pursuing a graduate degree with an NSF fellowship or becoming a teacher who will inspire new generations or working on cutting edge technology at an industry leader.
I used to worry that working with students on problems that interest them might be a distraction from my own research. I had, in earlier years, been asked by a local roboticist to help advise students on a project of his. I had no experience in robotics, but saw the excitement on the students’ faces. As the only faculty member positioned to co-advise, I knew that my saying “no” would crush their hopes. As I became more involved in their projects, helping to build a 3D printer from a kit, I became enamored with microcontrollers and the “maker” movement. At that point, it was just fun to build stuff and create an interactive project with a few lines of code; in my mind, it was completely decoupled from research. Then, two years ago, I began thinking about applying for an NSF grant, but was stumped as to what exciting research pathway I could propose. Serendipitously, a roboticist, whom I’d met through this robotics work, sent me a link to a TED talk with quadcopters cooperating to catch a ball in a net. A light bulb went off, and I saw a connection to the theoretical core of my research. I took a risk and proposed this robotics-based research program. To my delight and surprise, my proposal was funded!
This was not the only time a surprising connection came from working with a student. In fact, my first undergraduate researcher impacted my career in a way she may not even know. She sought me out one day as she was double-majoring in computer science and mathematics and had been told that my research straddled both. I felt completely unprepared as I had no list of specific problems; instead, I described my research on the fly, and my work in computational biology piqued her interest. She has since become a co-author with another undergraduate and two biochemists, but perhaps her most unexpected gift to me was a new collaboration, which I value deeply. As part of writing up her thesis, this student wanted to provide background on Lie groups, and she sought out the expertise of a mathematics professor. This professor saw a connection between the thesis work and the research area of another math professor. She encouraged us to start talking and thus began a collaboration for which I will be forever grateful.
As faculty at a research liberal arts institution, involving undergraduates in research is a core part of what we do. These budding researchers may not always be able to produce significant original contributions, but I can genuinely say my research path has been dramatically transformed for the better because of them. This is the biggest lesson I have learned: don’t underestimate the power of undergraduate researchers. They might directly contribute to your research, becoming co-authors on your next publication, or provide context and intuitions from things you’ve never thought about. And one day, those student interactions just might result in a connection that will transform who you are as a researcher. That connection could lead to a fantastic new collaborator or even a successful grant proposal. And, to top it all off, you get the amazing reward of knowing you played a small role in helping them in their own journey of discovering where to go next.