Choosing Ph.D. Research Topics
Throughout my twenty-ish years in academics (as an undergraduate and master's student at UVa; as staff at Caltech; and as a Ph.D. student at Claremont Graduate University), I've learned much about choosing topics of study through talking with researchers, observing the choices graduate students have made and the outcomes that resulted, and through reading books like The Craft of Research [1]. Now that I'm facing having to write papers and a dissertation of my own, I wanted to distill what I've learned into guidelines and write them down as reminders so that I don't lead myself astray over the coming years.
Ensure that your topic interests you intimately. This is probably the most important guideline I've learned. You're going to be living this topic for years. The Craft of Research emphasizes this with good reason. If you don't care about your topic, there's a good chance that you'll burn out and never finish your degree. Or you might finish and then abandon the field. I've seen both happen to dozens of people, including myself.
Choose a topic in which you want to be known as an expert, possibly for a good portion of your career. My advisor and other teachers at CGU have emphasized this, and as a student in his late thirties, attending an expensive school as a mechanism to change careers, I had already internalized this before this before starting graduate school. You're going to be living this topic for the years of your graduate work, and you're going to get to know it intimately. When you graduate and go for job interviews (academic or otherwise), you'll have a paper trail of publications on this topic, and your job talks will be on your dissertation. Ensure that the subject is relevant to the particulars of the career you want to pursue. For instance, if your career goals are to work in social networking technologies, don't spend your graduate career on wireless technology adoption.
Choose a topic that your advisor, your teachers or your fellow students can help you with. In a 2006 panel discussion, CGU faculty from varied fields emphasized that students who study topics their teachers have no familiarity with are at a disadvantage compared with students who choose topics within their teacher's areas of expertise. Similarly, they said that students who have fellow students with which they can discuss their research are better off than those who don't. Academic research areas are vast and take years to master. If you have no guidance, your research can be a long, lonely slog. There will be nobody to tell you how to start (by giving you a set of literature you should be familiar with) and guide you away from dead end paths, or paths that other people are already walking.
Choose the smallest possible problem to research that still satisfies your committee. This is also known as "the perfect dissertation is one that is finished quickly." Your primary goal as a Ph.D. student is to graduate quickly and get on with your life, not to get the Nobel Prize. You'll have plenty of time in the next thirty or more years to do that.
Tell a compelling story. Prof. Masakazu Konishi, who studies neuroethology at Caltech and is one of my former employers, emphasized this; The Craft of Research does also. Academic research is communication: it is meant to be shared and discussed. It's not just about learning or doing the things that interest you. If you expect to be successful at researching, and especially if you expect to be paid to do it, the story you tell must be interesting (compelling, even) and meaningful. At the very least, you want your field to think your story is compelling. Better yet is to have people outside your field care about what you do.
Tell a clean story. Another one from Prof. Konishi. Restrict your scope of your story so that it focuses on only what you want to talk about. Do this so that your audience doesn't get distracted by unanswered questions raised by barely related or broadly defined areas. From my graduate work at UVa in engineering, I learned to also restrict areas of variability that don't add to your story. Use clean, compelling logic to tie hypotheses to experimental design, design your experiments to give you clear results, and be able to tie your results clearly back to your hypotheses.
Your hypothesis should tell you something interesting no matter whether it is proved or disproved. One which Prof. Konishi was particularly emphatic about. You're going to be spending a lot of time and effort doing the research for your dissertation: make sure that time is well spent by ensuring that you will find something interesting to report. The worst thing that can happen to you as a researcher is to spend years of your life to arrive at inconclusive (and hence, un-publishable) results. I've seen more than a handful of graduate students abandon their topics and begin anew because they saw that the experiments they had been designing and doing were not going to be conclusive.
