One major idea here is to capture surplus value from transactions. In a typical school learning environment, the teacher teaches, the student learns, and very few lasting artefacts are produced -- even though generation after generation of student asks the same question over and over again.
In a research context, artefacts (e.g. papers) are produced. And in a free software context, similarly, code and documentation is produced. Free software even manages to gather value from the crowd, e.g. documentation is created through conversations in mailing lists.
My plan is to make a number of adaptations to PlanetMath that use existing tools to turn it into a "mash-up learning environment" for mathematics. The way people participate in each these dimensions will provide a stream of use data, including implicit and explicit data about motivation (including learning as an intrinsic motivation).
Furthermore, by focusing in particular on developing a "semantically adaptive personal learning environment", the smallest bits and pieces of "motivation" should come to the fore. For example, what does a student do when encountering a problem that is "too hard"? What kind of workflow surrounds writing new materials about a given subject? Ultimately, what makes learning fun?
Qualitative data in the form of interviews or similar will supplement these data flows.
part of Probation Report