In our project meeting yesterday we were discussing the concept of the "seeded" chats in which we provide problems for students to work on. The mechanism that we will have in place for delivering the problems is a link on the listing page. Users will come into the "lounge" and see a list of available problems and then, indented under those a list of rooms set up to discuss those. Next to the Problem name will be a link to an html file that we have defined when we enter the Problem into ConcertChat.
When thinking about this delivery process (which is all we have time to prepare before the launch of the new system), questions are raised about how problem solving sessions will be influenced by access to the problem beforehand.
In turn this connects to issues around the types of problems we use.
- Expected solutions problems: The plan so far has been to make the Current Problems of the Week available for free to students who agree to work on them in VMT chat sessions. While these are constructed response types of problems, there is almost always an answer or small set of answer that we expect to see. Would the ability to see the problem before the chat starts adversely affect the chat? We can imagine setting norms for the chats that encourage students to work together on solutions rather than deliver their answers. However, we are concerned in general with wanting to match the expectations and Internet orientations of many students which Wes has described as driven by "serendipity". Too many constraints may change the feel of the activity and thus the level of interest from students. There is probably little harm in models and suggestions about chat behaviors that are more likely to lead to good chats. In general, I think we'll accomplish more through the opt-in feedback whereby a group submits their session and requests feedback from us. Having asked, they are in a much better position to receive and make something out of such ideas.
Initially we could also try only posting enough about the problem to satisfy those who need to know what kind of problem it is before deciding to join a chat. If the numbers of chat sessions are small enough, we can manually deliver the url of the full problem when the group is ready.
- Open-ended problems: There may be less of an issue, or different issues for advanced viewing of the type of problem that we used in the pilot last May in which we presented a simple environment for students to explore and define for themselves the questions they then go on to investigate. It seems likely that this sort of problem would make sense as a group effort for most students at any time.
Those were the two main types. There were a number of variations we thought of: very difficult problems, create-a-something challenges, timed contests, etc. But they seemed to fall under one of those two general categories.
So, please contribute your thoughts about how we should present the PoWs in the context of VMT chats and whether we should offer open-ended problems as another problem type. You can only recommend doing open-end problems if you volunteer to create the problems. ; ) Remember also that there will be categories of problems other than "seeded". We will make categories where users can define the problems such as Study Group Problems and Any Kind of Problem.
We will make a decision next Weds., Aug. 31st. about the problem program for this Fall.