Analyzing Collaborative Websites

For several years now, classes at Georgia Tech have used a collaboration technology called CoWeb (Collaborative Websites) or Swikis. (See http://minnow.cc.gatech.edu/swiki for more information on the software.) The CoWeb allows students and teachers to collaborate in perhaps the easiest possible collaboration paradigm: Anyone can edit any page, and anyone can create new pages.

Though the tool has been in use in over a score of classes during this time, we've really only just started to understand its use -- and its NON-use. You can find some studies of these sites at http://coweb.cc.gatech.edu/csl/Papers. There are alot of questions which could be answered from even brief evaluations of these sites.

CoWebs have been used in CS, Architecture, Chemical Engineering, Biology, and elsewhere. How does use differ between domains? How does discussion differ in these different CoWebs? You could look at two or three of these and code the discussions for various content features (e.g., references to other classes, focus on homework vs. lecture, length of discussion.).
In CS classes, sometimes both newsgroups and CoWebs are used. How does discussion differ in each of these?
What is the pattern of discussion in a CoWeb, in any discipline? I've done some measururement of discussions in two different threaded discussion spaces (http://guzdial.cc.gatech.edu/papers/infoecol/). How is CoWeb discussion the same or different?