Over the weekend, edX announced some changes to the features in their discussion forums, and the changes illuminate some interesting issues in student behavior, pedagogy, and instructional design.
On the one hand, discussion forums are often seen as central to massive open online courses because in many courses it’s really the only space where students can speak back to professors or one another. If you believe that one of the key resources for learning in MOOCs is the community, then forums are often the central site for community interaction. At the same time though, discussion forums are often one of the most frustrating and difficult parts of a MOOC experience for students. It’s hard to have a functional conversation online with hundreds or thousands of people, and it’s hard to find what you are looking for in sprawling threads.
One of the fundamental problems with discussion forums is that the same technology is used to support several different types of conversations, two of which we might call “authoritative” and “discursive.” (Hat tip to Clay Shirky for this language, and some of the ideas herein). Authoritative conversations are those that are meant to converge on a single right answer, like the kinds of questions answered at StackOverflow. Discursive conversations are not meant to converge on a right answer; they are meant to allow people to explore ideas. Reddit is an archetypal space for these kinds of discussions. Typically, science, math and engineering courses are more interested in authoritative conversations, and humanities and professional classes are more interested in discursive conversations.
As a platform, edX has some built in biases towards the authoritative, which perhaps makes sense given it’s origins as a learning management system for teaching electrical engineering. If you look in the edX data documentation, you’ll find this nugget:
History: It used to be the case that Comment replies could nest much more deeply, but this was later capped at just these three levels (post, response, comment) much in the way that StackOverflow does.
That means that the discussion feature in edX used to allow infinite thread depth (the likes of which are often seen in Reddit AMA’s and other discursive forums), but the software was arbitrarily limited to support a particular vision of discussion.
When authoritative cultures and technologies are used in spaces where instructors want discursive conversations, the results can be quite messy. One terrible outcome is that teachers ask a discussion question and students answer it individually, one after another, without any reference to each others idea’s. Another problem is when experts who “know the answer” curtial the conversational space among novices. Imagine being in a course on moral reasoning and being told, “Share whatever ideas you have on this case study of deciding which of three people gets the last life jacket, just make sure that no one has ever made the point you are about to make.” Not surprisingly, courses in the humanities and professions have had more challenges with making discussion forums work than courses which work to help students understand a particular set of right answers.
With all this as backdrop, over the weekend edX added some features to distinguish between authoritative and discursive practices in the forums. From the announcement:
When discussion participants add a post, they will be able to identify the post as either asking a Question or starting a Discussion. Image
To distinguish between the types of posts, you will see a dialog bubble icon (Image) next to each Discussion post listed on the Discussion page. In the same list, each Question post will be identified with a question mark icon (Image) until it's answered, and a check mark icon after it's answered (Image). A new option for the list of posts (which we think the discussion staff will find particularly useful) is the ability to filter so that only unanswered questions appear.
So the changes that edX made were to try to signal between “questions” that are meant to be answered authoritatively and “discussions” that are meant to be pursued discursively. Questions can have right answers; discussions can’t. It will be interesting to see how these tools are used, if students and course teams can take advantage of the additional complexity, and if this resolves any of the challenges with forums.
More broadly, though, I think it illustrates an important larger point about technology-mediated learning, that all kinds of ideas, pedagogies, assumptions, and beliefs are baked into technical design decisions. This is one of the key issues that I hope we’ll explore in my class at Harvard this fall, T509-Massive: The Future of Learning at Scale. Learning technologies have a complex set of features, some of which are obvious and some of which are more subtle, and all of which can shape the way we learn.
For regular updates, follow me on Twitter at @bjfr and for my papers, presentations and so forth, visit EdTechResearcher.