[MyResearch] Ph.D chitchat with P. Salembier

Last friday, I met Pascal Salembier to have a brief chitchat about my phd project. I am pretty used to ask for some feedback from various persons (like Stefano about coordination theories or Thomas about psychology/methods/games). This time, my interest was put on epistemological concerns. Since we came up with this game idea to study how locative media modifies collaboration, I add few epistemological problems that I wanted to discuss: - cognitive ergonomics methodology - the use of game in HCI/CSCW/psychology - the outcome of the project: empirical knowledge versus applied guidelines: i would like to have both. - data analysis: I can use a program like actogram. And do I really need the video? maybe not because of the path I want to take. - the link between: paradigm/theoretical framework/research 'object'/methodologies

1. Cognitive ergonomics methodology

The canonical approach is: - first, ethnological method used to find properties, mechanisms implied in ecological situations. - second, when patterns are found, it's more a matter of having more precise questions, with an experimentl setting, with variables (like different type of interfaces, various display...) to get some insights and test them.

Conducting experiments without the first part could be detrimental, it's indeed possible that the test shows nothing...

Another way is like in the 'lutin' project: having an environment in which you can test stuff, do comparisons, see if things emerge... without any hardcore hypotheses. The point is not to have too much constraints, just few 'heuristical' hypotheses.

Outcome for my project: I am rather in the second approach in which I can use the game as a testbed to see how different groups use the tool and see if some patterns emerges. I still need few heuristics to analyse the wide load of data (soft hypotheses).

2. Use of games in HCI/CSCW

My concerns was: it seems fruitful to use games to study various phenomena in HCI (because participants are more motivated...) but is it OK epistemologically? The problem is that the task itself is not real, it is rather artificial (non ecological?).

It seems that it is not much of a problem since: - a game, especially a mobile computing one, involves participants in a real context (the physical world) with an ecological validity. A game in public space indeed create a certain kind of complexity.s - the domaine (task domain) is easier for both the participants and the experimenters (compared to air traffic control for instance). Besides, it is better to make students (my current would-be participants) doing a game than a really complex task they will never carry out. - participants would have a better implication than in a complex task

Ask Laurence Nigay, she seemed to have reviewed projects that used games in HCI.

Outcome for my project: it seems OK to use a mobile game.

3. Misc

The thesis needs an epistemological approach with: - a theoretical framework, there are various levels: from the more global one (distributed cognition, activity theory...) to much more precise thing (like the grounding theory from clark...). - well defined theoretical objects: what do I study? - well defined methods: how can I study these objects? - articulation between the objects and the methods: I need descriptive categories that don't come out from the blue. There must be a connection between them and the framework.

I should refine the research question so that I can set soft hypotheses to analyse the data and have a proper theoretical framework. A list of constraints would be great to find the task I want to use (like: I want collaboration, a task more complex than just a spatial one...)