In December 2008 I wrote a post related to choosing a dissertation topic.  More specifically, I was looking at weighing the choice between choosing a topic about which you are passionate, or choosing one that is marketable.   A few months went by and I went down the marketable path.  Then I essentially took a year off to teach myself the equivalent of a Masters in Psychology (may go get a real one after this is done…interesting stuff….).  I was also waiting for my chair to get back to me with news that he had gotten the data and needed my help managing it.  He didn’t, and I jumped to the conclusion that he hadn’t gotten the data.

Throughout this last year I have collected articles that I saw that seemed related to the topic I was planning to pursue.  Or so I thought.  When I went through those articles this weekend I discovered that they were scattered; it was nearly impossible to identify the theme of those articles (which, mind you, I had not read).

This weekend I reflected on that lack of coherence.  Part of the problem was lack of time; I picked up anything that seemed passingly related and set it aside for later.  But part was a lack of clarity around the core question I was trying to ask.

You see, a dissertation is a research project, meaning that you have to have a question or hypothesis to test.  And to me, it’s even MORE important to know that your data will be able to answer your question.  You don’t need to know the answer right now, and in fact you shouldn’t.  But without a clear question and confidence that the answer is in there, it becomes hard to focus your research and writing.

For someone intending to gather their own data, this is a bit different.  The question for you is whether you can get the right sample, perform the correct techniques, generate unbiased  instruments  for use with human subjects, and in the end get the pieces you need to answer your question.

But part of my area of expertise is large data sets, and it is impractical for a dissertation project to collect one.  It is also unnecessary.  The National Center for Education Statistics (NCES) does large scale surveys/studies and makes much of the data publicly available.  As a great professor told me when I just started my PhD, those data sets contain thousands of unwritten dissertations available for the taking.  And that is WITHOUT any potential data from my advisor.

Anyway, the point here is that I realized my inability to focus was related to having no idea what data I would have.  I was trying to cover every possible question my advisors data might hold because I didn’t know what was there.  Rather than continue to thrash like that, I have decided to work off of one of the NCES data sets for a similar topic.

There are countless great ideas for dissertations but it is critical that, while assessing your ideas, you consider whether they are actually feasible.  The dissertation should be the worst research project you ever do, but you need to make sure that you can finish it.  That means a tight topic, limited scope, and making sure that the data you are either collecting or using can actually answer the question you are asking.