Somehow, shockingly, I’ve arrived at the point where I’m just a few mere months from finishing my coursework for my doctoral program (okay, 50 days, but who’s counting?), which means that next semester, I get down to the business of starting my dissertation. One of the interesting things about being in a highly interdisciplinary program like mine is that your dissertation research can be a lot of things. It can be qualitative, it can be quantitative. It can be rigorously scientific and data-driven or it can be squishy and social science-y (perhaps I’m betraying some of my biases here in these descriptions).
If it weren’t enough that I had so many endless options available to me, this semester I’m taking two classes that couldn’t be more different in terms of methodology. One is a data collection class from the Survey Methodology department. We complete homework assignments in which we calculate response and cooperation rates for surveys, determining disposition for 20 different categories of response/non-response/deferral, and deciding which response and cooperation rate formula is most appropriate for this sample. My other class is a qualitative methods class in the communications department. On the first day of that class, I uncomfortably took down the notes “qual methods: implies multiple truths, not one TRUTH – people have different meaning.”
I count myself lucky to be in a discipline in which I have so many methodological tools in my belt, rather than rely on one method to answer all my questions. But then again, how do I choose which tool to pull out of the belt when faced with a problem, like having to write a dissertation?
I came into my doctoral program with a pretty clear idea of the problem I wanted to address – assessing the value of shared data and somehow quantifying reuse. I envisioned my solution involving some sort of machine learning algorithm that would try to predict usefulness of datasets (because HOW COOL WOULD THAT BE?). Then, halfway through the program, my awesome advisor moved to a new university, and I moved to a new advisor who was equally awesome but seemed to have much more of a qualitative approach. I got very excited about these methods, which were really new to me, and started applying them to a new problem that was also very close to my heart – scientific hackathons, which I’ve been closely involved with for several years. This kind of approach would necessitate an almost entirely qualitative approach – I’d be doing ethnographic observation, in-depth interviews, and so on.
So now, here I find myself 50 days away from the big choice. What’s my dissertation topic? The thing I like to keep in mind is that this doesn’t necessarily mean ALL that much in the long run. This isn’t the sum of my life’s work. It’s one of many large research projects I’ll undertake. Still, I want it to be something that’s meaningful and worthwhile and personally rewarding. And perhaps most importantly of all, I want to use a methodology that makes me feel comfortable. Do I want to talk to people about their truth? I’ve learned some unexpected things using those methodologies and I’m glad I’ve learned something about how to do that kind of research, but in the end, I don’t think I want to be a qual researcher. I want numbers, data, hard facts.
I guess I really knew this was what I would end up deciding in the second or third week of my qual methods class. The professor asked a question about how one might interpret some type of qualitative data, and I answered with a response along the lines of “well, you could verify the responses by cross-checking against existing, verified datasets of a similar population.” She gave me a very odd look, and paused, seemingly uncertain how to respond to this strange alien in her class, and then responded, “You ARE very quantitative, aren’t you?”