How to write a thesis

Writing a thesis is a difficult undertaking. Before I started writing mine, I hadn’t written any assignment longer than 30 pages (my Bachelor’s essay), and it was quite a step up from that to having to structure a huge complex of data that I’d gathered on my own, analyze it and bring it together in a coherent academic argument.

Luckily, I was well helped along the way by my supervisor, Morten, who really reeled me in from time to time when I was going off in weird and unsustainable directions, which happened fairly regularly. He gave me a lot of pointers, which I have summed up here for anybody about to write a major piece of academic argumentation. It may seem simple enough, but trust me: Once you get involved in it, you lose yourself to the writing, and it is difficult to avoid being overly esoteric with regards to your special niche of interest.

  • Be overly pedagogical! Keep a continuous meta-discourse going to explain to the reader why this bit of information is relevant in the grand scheme of things. It may seem obvious to you, since you know what is coming. But the unaccustomed reader won’t.
  • Use lots of part conclusions! Sum up again and again how each bit of analysis is relevant and necessary to make sense of your overall argument.
  • Focus on readability! Don’t use more than a handful abbreviations that you can reasonably expect the reader to know in advance. Use clear examples to explain difficult terms and processes!
  • Be very careful with descriptive passages. It can easily become either dry or boring or light-weight and irrelevant. Keep your focus on the relevant scientific observations. Those are the ones that you are meant to pursue!
  • Make it perfectly clear to yourself which academic or scientific tradition you aiming to be part of. Are you going for the anthropological insights, or the psychological qualities, or perhaps the computer science bits? There’s no way you can appeal to all, and your thesis will suffer from lack of focus, otherwise.
  • Be analytical: Use quotes or specific data to underline your analyses and conclusions. Hack the data! Fashion surprising and worthwhile points from your empirical descriptions.
  • Write descriptively in order to support your analysis – but don’t write naïvely. The description can be an analysis in its own right if used to expose analytically interesting situations and issues.
  • Use and express clear levels within the text: Who is saying what? When are you being analytical and when are you being descriptive? Use meta-commentary to separate the two, but don’t be judgmental. Try colouring the text so that you can see where you are analyzing and when you are describing. Keep these in separate sections! Otherwise it will confuse the reader!
  • Make clear distinctions between what your informants are saying, and what you are saying: Are you using their metaphors and terms? When are you speaking and when are they speaking? You cannot be reflective and critical when using their terms. Use italics and quotes to signify that you are aware of the difference!
  • Be reflective all the time: Ask: Which implicit assumptions do your informants have that shape their demeanor and convictions? For example: What assumptions are inherent in the idea of the transparency of a computer program? How does this assumption shape relations between people?
  • Focus on the relations between your informants! What does it mean to be part of this group? Is it a group? Where do their shared bonds lie?
  • Pick a theoretical perspective and give it more depth! Illuminate it from different angles through various analytical means. Dig deeper!
  • Use diagrams to illustrate and explain tricky analytical points that you find central. Often, a good diagram will express a thousand words of analysis.
  • Each chapter of the thesis should be a paper in its own right – containing its own analytical focus and conclusion. But at the same time, it should lead on to the next chapter. Ask yourself: How does this chapter lead on to what I discuss in the next chapter? Is there a feeling of natural flow between the analyses?
  • Layout the text as it has to be in the final version. It will make it a lot easier for you to see if you are within the formal word and page limit. Writing too much will require rewriting and cutting, which is arduous and difficult! Better to write it right the first time.
  • Have a draft chapter ready for review for every meeting with your supervisor. Write a letter along with the draft: Describe how the draft fits with the greater whole of the thesis – what function it fulfills. Make it easy for the supervisor to comment it in a way that can help you!

Well, I’m sure this seems like pretty self-evident advice, but it is still hard to remember when you’re getting carried away writing about your very favourite obscure detail about the history of the Unix operating system. And you know it has to go, the moment you finished it…


  1. Pingback: Advice on writing a thesis « figuring out computer science
  2. Pingback: How to write a thesis « Dreamwalkn in Vermont
  3. Lasse Havelund

    Layout the text as it has to be in the final version. It will make it a lot easier for you to see if you are within the formal word and page limit. Writing too much will require rewriting and cutting, which is arduous and difficult! Better to write it right the first time.

    Just use LaTeX from the beginning ;). Thanks for taking the time to compile this list; despite still being in the gymnasium, I can see some of these points being relevant for some of my reports and essays. And congratulations on finishing your thesis!

Post a comment

You may use the following HTML:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>