Getting a Sense of What Worked (and What Didn't... and What Worked Differently than We Expected)
Once the students were comfortable in their roles, the courses converged on the shared forum space. Over the four weeks of shared forum discussion, students read the same materials and each class was given the opportunity to perform a different role each week. Instead of a handful of launch posts, however, there were now twenty choices for response and at least sixty posts to expand a discussion. The students now had to make complex choices about how they would respond to and interact with one another. Crafting an effective launch post became important as the other roles seemed to respond to launch posts very strategically.
Usually, the launch post with the most interesting or well-articulated question in the title garnered the most responses. Furthermore, we witnessed a clear snowball effect benefiting the early movers: Posts put up early often got responses, and posts with several replies usually generated even more replies. Perhaps we should have anticipated this outcome, but we did not think student responses would be so closely tied to an ethos of expediancy. But that's not to say that this ethos wasn't in some ways sophisticated. Students learned that crafting a launch post title that included a question typically gathered more responses (launches with questions averaged 3.6 responses while those without questions averaged only 2.26 responses).
Students who posted launches after the deadline were usually excluded from a conversation that was moving on without them. Similarly, students whose launch posts were either too complex or too off topic usually saw fewer or no responses (though as instructors we often introduced the complex posts in classes).
Unfortunately, due to the death of the server hosting our forum, we cannot provide links to the project in its entirety. In the following sections, however, we do recreate specific posts and discussions for which we have secured the authors' permissions.