Thesis supervision

aje on November 2, 2015

This page provides general advice for students writing a thesis.

The most important first step is to identify a research topic. I consider the power of economic reasoning to stem from its applicability, and take a broad and eclectic position of what would constitute suitable subject material. For a general management thesis I don’t require students to work on the same research topics that I do.

A successful thesis will then accomplish three things:

(1) Choose an insightful research question

The main difference between a very good thesis and an excellent thesis is whether or not you articulate, and answer, a good research question. In most theses that I see, this isn’t the case. Typically students will identify an interesting topic, and then proceed to investigate it. But the purpose of a thesis isn’t for you to learn about something, it’s about contributing to our collective understanding. I don’t expect students to have a good research question at the beginning of their project, but be wary of reaching the end of it without having one.

(2) Provide a rigorous literature review

A literature review is more than just a discussion of your topic, it is supposed to survey the existing literature that relates to your research question. For more see here:

(3) Utilise the right methodological framework

To start off with, I recommend the following articles on research design:

If you use standard research methods then be aware of the social priming replication crisis, publication bias and avoid p hacking. For more listen to:

I would expect all theses to fit into one of the following categories:

      • Conceptual – this could be theory based, a comprehensive literature review, or some other form of non-empirical paper
      • Empirical – this is more common and investigates/answers the research question utilising data. We can split an empirical paper into 4 main sub-types:
        • Quantitative – this is the typical format for an economics paper and is based on the use of statistical tests and model building to analyse measurable phenomena.
          • The quality of a model rests on the use of control variables
          • The quality of a survey rests on the randomization strategy
            • Note that the common problems with surveys are survivor bias; the fact that they capture a snapshot rather than a trend; they provide only an indirect view of behaviour; and they have the potential for dishonesty from recipients. Useful remedies include posing questions in hypothetical situations, and a recognition/discussion that you are establishing upper or lower bounds (for more on the use of surveys in hostile contexts see Frye, 2017, p. 33-35)
        • Qualitative – these are descriptive accounts that value an interpretive approach
        • Mixed methods – i.e. combinations of both quantitative and qualitative research methods
        • Hybrid methods – e.g. the comparative method (QCA). (See here for an introduction to the comparative method).

Perhaps the key ingredient to a successful thesis is to demonstrate competent project planning.

This is crucial because it determines whether the experience is enjoyable or not. The following are necessary (but not sufficient) characteristics you need to have:

  • Enthusiasm for the research question (and not just the research topic)
  • Genuine desire to have people read your work
  • Ability to self-motivate
  • Swift communication

I will either provide you with detailed feedback on a full draft, or brief feedback on specific questions, but you should not expect me to provide multiple rounds of comments throughout the process. Depending on how many students I supervise in any given year, I intend to provide a similar amount of help to each and will be unable to devote significant time to your project close to the deadline. 

When planning the writing of the thesis take a look at:

This is also useful: Baylor University research planner guide. And Barry Weingast’s ‘Caltech Rules for Writing Papers‘.

If you get to present your work, here’s a good guide for creating a poster (and here). Don’t forget to include a clear plastic wallet with printed copies, and one for business cards.

Grading

Here is the way in which I judge academic articles and conference presentations. It contains information relating to research articles; theses; the use of data; and sociology challenges:

For more details on the grade ranges that I typically employ see page 7 of my guide for students, however you should adjust the passing grades such that what I deem to be a C grade for a thesis would get a mark of 55-60; a B is 70-80 and an A is 85+. These are only general guidelines and there’ll always be a gap between my judgement and your understanding of my judgment. But just because the grading is subjective does not make it arbitrary.

Finally, if you’re interested in a career in academic economics, here is advice on surviving grad school. And if you are a female student, do take a look at this “Women in Economics” video series.

Last updated: September 2022