Jesse Lawson

Software engineering, artificial intelligence, writing, and open-source tools

Dec 14, 2015

Researching areas without a lot of research

For a doctoral seminar I am in, one of the requirements is to compile a specific number of references in support of my research topic. Since my topic centers around how predictors of student behavior variables have not changed since we started scrutinously looking at them in the 60s and 70s – and whether we are learning new insights about student behavior predictors with the use of machine learning and nother advance statistical methods – you can guess that the studies directly related to my field of study are few and far between.

So here I am in a doctoral class, a doctoral student and all, and I’m told that I need to find X number of studies that specifically address machine learning in higher education. This is not a small number, by the way. We’re talking well over a few dozen. For a typical psychological paper, I’m okay with hundreds of references directly relating to some facet of the research I am presenting, but my field is a sorely understudied field in higher education scholarship. What I am getting to by this is that the scholarship on this issue is sparse – so sparse, in fact, that I have written the only book dedicated specifically to machine learning in higher education.

Perhaps this is the typical doctoral student egotism coming out, but I think once you surpass the master’s level training there is a point in your doctoral studies that you become somewhat of an expert in your field. Eventually, your knowledge surpasses your faculty’s ability to contribute constructively to your argument, and the efficacy of their input is reduced to APA style guidelines and (hopefully) rhetorical structure of your dissertation components. When you are presented with an assignment that has specific boundaries and your research topic is sparsely covered, the only way to ensure that you meet this arbitrary study limit is to expand your field of vision into other types of studies that may relate tangentially to your topic. Am I crazy, here?

Well this is not possible with this particular faculty person, who is convinced after a short run in the university library that the word “machine learning” definitely brings back “more than a thousand” results. At some point, I think the frustration I am dealing with – a frustration that is no doubt shared by all doctoral students at one point or another – leads to poor coping behavior.

I hope that when I start teaching graduate students I work more on guiding their own research than I do on treating them like they just fell of the turnip truck. Interestingly, this is a common criticism of doctoral students these days (and perhaps all days… I can’t know for sure). In my master’s program I was treated as a professional and at least a junior expert bringing to a discussion some knowledge about a complex subject. In my doctoral seminars, though, it doesn’t matter how many classes I have taught, books I have written and edited, and consulting work in this very field I have done. There is an air of “you are a student, I am a teacher” that I really, really don’t like.

Unfortunately, I am less than 10% done as of this fine Monday morning. Here’s to persistence and commitment.