Team from Purdue Calumet talking about their student-driven approach to Second Life. Students earn experiential credit by coming into Second Life. Group has faced typical problems of funding, development, sustainability.
Wanting to measure affective state of learning: accuracy of affective state detection by facial expressions is 70% accurate. Using single-source camera to gather data; doing pretty sophisticated analysis of facial expressions (Gabor Features?). Presentation stayed pretty technical--not immediately relevant, but interesting. One presenter laid out nicely how research in VW needs to employ mixed methods: qualitative research on learning experience, quantitative research on technology (performance, stability, etc.).
Mark Mabrito in English at Calument was not present, but he is teaching a class called "Designing Virtual Worlds" -- developing a certificate; part of an online certificate in writing for interactive media.