SIKS – Research Methodology

Last week, I partook in a three-day course on research methodology (together with Thomas) at Landgoed Huize Bergen in Vught, organised by the SIKS research school.

To get directly to the point, I found the practical assignment Hans Weigand gave us the most useful part of the course. Of course, this assignment was introduced via a short lecture, in which Hans explained the concepts of the problem bundle, the research model and the research framework. What it all boiled down to was that you should start by placing your research subject into context and draw causal links between concepts related to your research. For example, looking at the domain of serious games, the phenomenon that training is expensive is causally related to the amount of training people get, which in turn results in lower performance. Drawing a graph with this information proved to be insightful. From this graph, we designed a rough outline of which steps to take during our PhD trajectory.

Roel Wieringa‘s talk was quite a bit more abstract while Hans Akkermans‘ discourse was a lot more down-to-earth and concrete. The latter gave some very useful examples of actual research methodologies, such as types of argumentation, validation and observation. There were also some talks by invited speakers that were domain-specific, viz. by John-Jules Meyer, about the philosophy of logic (very interesting! but not that relevant to the course); Antal van den Bosch about machine learning; Djoerd Hiemstra on information retrieval (he made a good point about that you should use research methods usually used in your field but that you should not strictly cling to them); and a talk by Virginia Dignum, about multi-agent systems (interesting, but very general and more like an introduction to the field of MAS). It was nice to see the various topics, but to me it seemed that only Djoerd had something to say about the how (the methodology) instead of only the what (the research subject itself).

The same remark holds for the invited PhD students (Evgeny Knutov and Rory Sie), who gave a talk about their PhD research, but mostly focused on the subject and not on the methodology.

All in all, the course was at times useful, at other times interesting (mostly in combination), but now and then vague or unrelated to the overall goal. At least I have a clearer picture of my research and an interesting idea for an adaptable poster, in cooperation with Thomas, based on Hans Weigand’s assignment—time will tell how this will turn out.

Fig. 1: The group of attendees at the course (I can be found in the back row, to the left).