Be always sure you’re right, then go ahead.
Davy Crockett
After advising a couple of bachelor and master theses in media and computer science, I’ve created a short document to help students with planning their thesis topic. As I’ve written (and spoken) about in multiple postings (and videos), a thesis is a complex issue. You need more than a good topic, or motivation, or skills. It has to be — at least — good enough on a multitude of variables.
As a metaphor, just think about it like a couple of pillars standing in a water — you don’t want to drown or even touch the water. So make sure that each pillar is above the water line.
As bare minimum, if you decide for a topic you should have the combination of thematic interest x technical competence x competent advisor. You need to be interested in the topic to continuously work for about 6 months, you need to have the skills to actually implement it (sketches are easy), and you need to get feedback along the way by someone who gives a fuck and knows the field.
So, as a first step, it pays to just list the possible topics (what’s interesting to you), your competences (technical or otherwise), and which advisor has which knowledge skills.
That should both stimulate possible topics and narrow down the list as well.
If you have a couple of likely topics, it’s time to do some thought ‘experiments’ — just see whether one of those ideas actually does hold the weight of six months of work. And the stress of a grade that might be somewhat relevant.
So, here’s a table that might be helpful (click on it to enlarge it):
The columns are the steps (chapters) that are part of the typical bachelor and master thesis at the university I work for. Think of these columns as stepping stones to a good product via a human-centered design process. Thing is, you must be able to deliver great work in each of these steps. If one of them is weak, it does become the weak link that will break your work.
For example:
- If you cannot see the use in the product — why should I care — then why should anyone care about what you do? (Why should you, for that matter?)
- If you do not have access to a good literature basis, the analysis lacks the foundation. You can do interviews and surveys, but that will take a lot of time without a good literature basis.
- If you do not have access to people who evaluate your work formatively, how can you expect to improve it?
- If you do not have the skills to actually implement the project idea, how can you actually evaluate the product? (Yeah, you can use high-fidelity prototypes, but still, there’s a difference, if only that future employers know you can not only talk the talk but also walk the walk.)
So, it pays to play around with the table for a while. Just fill it in, then decide. It does work better if you see all the options, and both the stepping stones and the sink holes.
BTW, the German document is here (.docx, .pdf), the English one here (.docx, .pdf).
As usual, constructive feedback greatly appreciated. 🙂