Oxy’s ten-week summer research program concluded last Wednesday with student presentations. I was traveling to a conference and was unfortunately unable to attend, but I met another professor at a small liberal arts college at that conference, and we shared our respective experiences mentoring undergraduate research. I just wanted to share some assorted thoughts.
I am a goal-oriented person, so my first thought is about the goal of undergraduate research. I think there are two clear goals for me. The first goal is to provide students with research experience, and thereby developing their overall skills as an academic; the second goal is the actual goal of research, to discover ideas with scientific value (and maybe publish a paper on it). It is unsurprising that this reflects the “teaching” and “research” sides of being a faculty member as well.
I originally wanted to write about the process of recruiting research students, and it took me several hours while writing this post to understand the constraints that the two goals impose. Although my initial scapegoat is Oxy’s small population of computer science students, I can replace “computer science students” with “insert-disciplines-here students” and it will be just as true. I think when I try to find good students for projects (or good projects for students), there are two factors at play:
- Does the student have the skills and ability to work on the project? For me, this is often some basic knowledge of programming, although it may not be necessary for some projects. This question hides a lot of complexity about the development of the student – being able to start on the project is the minimum requirement for improving those same skills and building their interest as potential future researchers.
- Will the project lead to scientifically valuable results? Especially for me as untenured faculty, I would prefer not to spend time on research projects that do not lead to useful results. (Strangely, I’m okay with leading independent studies, but I judge those very differently from directed research.) The results do not have to be directly publishable, but I (or another student) should be able to pick it up and carry it forward.
It seems obvious, but the difficulty of finding students comes from the fact that very few scientifically interesting questions are pursuable with the skills of an undergraduate. It is not a simple issue of looking through the list of interested students, the list of interesting projects, then picking one from each. Instead, I must consider the skills of the student, and see if any project has a semi-independent chunk that would fit that skill set. Since each project can be broken down in multiple ways, the student-project matchmaking becomes non-trivial. I am partial to a more interactive approach and prefer leaving the final deliverable, or even the scope, of a project undefined until the middle of the semester/summer.
Even knowing the rough direction of a project, I have yet to master the art of guiding students through the research process. I am averse to projects where students just analyze data from experiments, and unlike many physical sciences, few computer science experiments which require human babysitting. Instead, most of the time I have my students explore the literature and generate ideas, as I help them understand the basic concepts and direct their attention to topics that interest me. I must admit, however, that I struggle to balance giving students with enough information without overwhelming them. More than one student have suggested that doing research with me can be frustrating, and I’ve never figured out why.
There may be several factors at work here. One is the illusion of transparency – I will explain an abstract concept and see my students nod, but they will have trouble applying that concept in their assignments. Unobvious confusion is not a problem limited to my research, as I felt the same way when teaching loops in CS1. In that specific case, I provided worked examples that also described my step-by-step reasoning, and a similar solution may work for research. I’ve also decided that although weekly meetings are sufficient to track progress, I need semi-weekly check-ins to make sure that students understood what they were supposed to do. The more frequent meeting schedule prevents wasted effort, and also gives me more immediate feedback on whether the student needs help.
The other factors of student frustration are less clear to me. One may be how most research students work alone. I must distinguish between working on a project alone but spending time with students on other projects, from working alone and not talking to anyone at all. Other faculty have anecdotally suggested that I should avoid both types of isolation, although personally I have had success with students who worked on their projects by themselves. I suspect that the summer schedule also has an effect – compared to the regular classes and office hours during the school year, the one meeting a week may give the impression that I am unavailable at other times. I considered organizing work sessions to increase face-to-face time, but missed the opportunity to do so at the beginning of the summer.
A third potential factor comes from asking whether I interact differently as a research mentor (as opposed to a teacher). One difference may be higher expectations for students to figure things out on their own. This is especially during the summer when they are not taking classes and are (in theory) doing research full time. I don’t have data on how many hours my students work, nor do I have a target number of hours for how much they should work. It is at least possible that I am being unrealistic in my expectations, and I should informally survey my students about this.
I’ve been fairly critical in this post, but I do enjoy working with students on research. Besides forcing me to keep up with the literature, research focuses my attention on the students individually, and often pushes them in ways that normal classes cannot. I think many students who decide to do research see it as this grand pursuit of the Truth, and if nothing else, I think I can use that reminder more often.