This afternoon, at the lab, we had a discussion about methodologies. The goal was not to agree on a common methology for all our projects, but rather to have a discussion of what we can use and in which context. One of the issue that lead to this was the fact that we know what we're not using (pure experimental research paradigm for instance) but it's harder to define what we should use. Especially given that our research projects have different purposes ranging from studying people's behavior when collaborating too designing applications or services that meet specfific needs. Jean-Baptiste presented what is called "Design-Based research" in educational science because he thought this might be of interest in terms of mixing qualitative and quantitative methods in an 'interventionist' way. Hence we had a discussion about it and commonly agreed that some project can use it (design-oriented projects) and some cannot (when we study behavior rather than designing artefacts). Pierre was still wondering about referring to this theory in our research.

Then Pierre (lab director) summarized the main characteristics of what our lab should put the emphasis on:

intervention or design-based approach.
variation: studying variation, to compare things (but not necessarily having controlled group, it's rather a matter of comparing different interfaces than comparing a group with an artefact to a group without it).
investigating the PROCESSES rather than the outcomes. By processes meant the interaction with the artefacts/application/service but also the group processes when they collaborate using it. The outcome or the performance is often less intersting than the processes that occured.
we collect data (be it quantitative or qualitative) to analyze what happened.
granularity of our research: emphasis on small tests/experiment/investigation (which - in addition - are not pre-experiment to fix bugs or tune settings)
generalization: how to augment the external validity of our results? by having larger scale investigations or doing them across contexts (task/people/settings)? or do we neeed generalizations.

The discussion was quite messy but it was really relevant for us to talk about this. Of course, it's still in process. Sometimes, in the discussion, I was not so comfortable by all of this (we need to set boundaries between research goals and the corresponding methodologies); but anyway, that's what happened when people with different perspectives talk with each other. Actually, even though we want to mix methods, it's clear that the main paradigm at CRAFT is positivist and bound to quantitative methods. Then, it's more a matter of discussing how to integrate other methods (qualitative) in a proper way.