Research question checklist

Author: Stephen J. Turnbull
Organization: Faculty of Engineering, Information, and Systems at the University of Tsukuba
Contact: Stephen J. Turnbull <turnbull@sk.tsukuba.ac.jp>
Date: June 19,2020
Copyright: 2020, Stephen J. Turnbull
Topic:Research

A research question is a specific problem in understanding the topic you are researching. Unlike the research topic, it will be stated in technical terms related to the specific field and method of research. In a theoretical work, it might be stated as a proposition to be proved. In empirical research, it might be stated as a statistical model and a hypothesis test. The topic statement is intended to excite interest and draw an audience. Research questions express accurately what you are going to try to prove and often indicate how. Research questions are only meaningful when embedded in a model, while the topic statement is usually embedded in the general knowledge of the audience.

As a matter of research strategy, you should have many research questions, more than you can possibly address. Until you have a set of results that is sufficient for your research goal, you should keep adding more questions. There are for reasons for this. First, research is by definition uncertain. You may not get reportable results. If you have only a short list, you're stuck. Second, your leftover research questions help to define your work's relationship to the field. Third, they are often useful to present in your discussion section. Finally, the most research questions you come up with, the better you are prepared when an examiner asks, "why didn't you do X?"

For each of your research questions, you should confirm the following:

  1. How does answering this question advance my research?
  2. Is it clear what model the question refers to? For example, in a statistical hypothesis test on a regression model, whether the disturbance term is i.i.d. or not needs to be clear.
  3. Is it clear which parts of the model are tested in the question?
  4. Is the methodology clear?
  5. Is the criterion for a "successful" answer clear? (For example, in a mathematical theory, you either prove the theorem or not. In a statistical hypothesis test, the level of significance and power are the criteria.)
  6. What result did you expect before doing the research?
  7. What rationale do you have for the expected result?