Finding One’s Way toward the Dissertation

By | November 18, 2013

Every doctoral student has moments of discouragement and occasional feelings of being lost in the process. The important part is not to let these moments become long-term or permanent. It’s not unusual for people to back up several times during the course of the process, rethink where they’re going and what they want to say and how to find out about it, and emerge with a much stronger product.

For the first time in several years, I am once again working with beginning PhD students. It’s interesting how fast we forget what a confusing and difficult time that is in a new scholar’s life. In retrospect, it’s perfectly clear how we managed to work things through for ourselves, and how all the students who worked with us moved along so smoothly and efficiently through the maze. But all retrospection is misleading at best, and the directions which seem so self-evident to us when we’re all done were never so clear at the time. It’s very easy to forget how many times we may have had to change gears and sometimes completely change the entire transmission.

I am a good example of this. About a year and a half into the process, I had to completely rethink what my dissertation was going to be about and even what my methods were going to be. My topic was the implementation of technological innovation in organizations, specifically in my case a variety of computer systems in local government and quasi-government agencies. Originally, I was planning an analysis of quantitative data that Ev Rogers and I had collected during a series of site visits around the country. But when in April I was told by my department that my appointment to an assistant professorship in September depended on my having finished my dissertation, I began to panic. I realized that the data we had were not really reliable enough to support the kind of analysis I wanted to do. This isn’t all that uncommon in research studies. Often it’s not until you finish collecting the data that you realize the questions you should have asked in the first place. But when your dissertation depends on this, it’s panic time.

At that point, I began to rethink how I would approach the problem. I found the key in a diagram of innovation processes that I had dashed off during one of our site visits while trying to make sense of some of the data we collected up to that point. That diagram actually constituted a process model that could be tested in a qualitative comparative case analysis. Then I realized that since well over a year had passed since the data were originally collected, if I could collect some new data about the same cases I could turn it into a longitudinal study with two time points – really necessary for testing a process model. Looking over our previous cases, it seemed that Baltimore would be an ideal site for me. We had some seven separate cases in the Baltimore area, plus I really enjoyed Baltimore as a city.

So I arranged to spend a week in Baltimore re-interviewing most of the people I interviewed a year before, discussing what had happened in their particular cases since the last interviews. I was then able to triangulate the different activities relating to the computer system in question in each of the seven cases, and demonstrate a high degree of convergence between the actual behavior in these agencies and what my model would predict. I finished the interviews and analysis sometime in early August, and then wrote the final draft of my dissertation in some 10 days. Needless to say, I was not very good company during that period, but I got it done – and even more, I and my committee were both pleased with it. And even even more, the model has held up remarkably well, despite enormous changes in the technology itself.

I’m relating all this not to toot my own horn about what a great dissertation I wrote – simply to point out that running into a brick wall in the course of the dissertation can be surmounted, either by dismantling the bricks or by finding a path around the wall. It may take a little time and some inspiration, but it can come if you really want it to. If you succumb to discouragement at this relatively early stage, then you will find it very difficult to complete the process. What’s needed is that you be interested enough in your topic that it will sustain you at least until you finish.

I’ll have some more comments on this topic in a day or two.

  • This is great stuff!

  • One thing that can be generalized from you experience is that frequently dissertation writers need to scale back the project, and that applies even when there’s no data set etc.

    • DrEvel1

      Absolutely correct, Rega. The issue is conceptual and (dare I say it?) philosophical, and the problem is the quantity and complexity of the ideas. Empirical data and/or mathematical operations are simply ideas represented in alternative languages, even as, say, lex parsimoniae and ding an sich are ideas expressed in Latin and German, respectively.