Bill's Research Idea

1. For next quarter, it might be a good idea to spend time in the seminar looking at the data we have gathered this quarter and trying to apply various descriptive and analytical methods to them. We will not be offering the Pixels course next quarter and will devote all of our time to data analysis and getting stuff written up. Extra eyes, ears and brains of seminar attendees could help us coinsiderably.
2. What we have looked at this quarter has been both inspiring and frustrating. I took comfort from the fact that there are ways to describe the somewhat complex data we have mathematically. But I didn't think that what we read (Bayesian nets, HMMs) can do all of what we need to do.
3. In that connection, I rather liked Daryl's idea about having some students work to make some simple tools that would help us with 2. It would also give them projects to work on that have real and immediate use.
4. Finally, we need to remain focused on educational applications. There are things that I am reading in systems dynamics and cognitive neuroscience that might be good to share that might tie cognitive (and biological) theories of learning to computational methods by which to implement them.