Social shift in MOOCs

Executive summary: Matti has been looking some MOOC related research taking place in UC Berkeley. One emerging theme is adding peer-to-peer interaction into the exercises to support learning. Matti suggests that this might be a weak signal and learning platform developers ought to have thought on this.

The big promise of massive open online courses (MOOCs) has been opportunity to access to materials from the worlds leading universities and gain knowledge from there. Education that scales is easy to access and follow and allows you to familiarize different topics from the comfort of your own computer.

Currently most MOOCs (based on my experience on the edX and Coursera platforms, at least) seem to have rather similar structure: there is a video, some slides and talking head of the teacher and quizzes. There are assignments where peer-feedback is used to scale the mechanisms, as the computational machine is not yet advanced enough to work there.

However I do think that there is a change coming into this field: it is not just the access to the material that creates value, rather it is the interaction between participants to explore, reshape and co-create knowledge that is vital in the learning.

My suggestion of the social shift is more than just the peer-feedback or forums; it is integrating peer-discussion or help into the educational platform. Here are some examples of what I’m seeing taking place in the MOOC research field

• Initial Experiences with Small Group Discussions in MOOCs is early work conducted at UC Berkeley on integrating a chat discussion as part of choosing an answer in a multichoice question. Students experienced that discussing an answer with their peers required them to explain their thinking more and made the answers more solid.
• Talkabout: Small-group Discussions in Massive Global Classes studied scheduled discussion sessions in a MOOC environment, done together by researchers from Stanfoed and UC San Diego. Students found these discussions valuable, however the choosing these times before the session caused students often to forget these sessions.
• Remote pair programming (RPP) in massively open online courses (MOOCs) is work in progress to enhance programming learning experience by introducing pair programming (=two coders, one desktop, one code) into MOOCs. The pair programming approach has been promoted in software industry as it creases opportunities to peer-learning and peer-debugging.

Let us assume that this is not just a buzz caused by my way of looking the world as social thing, but that these developments are the next steps in MOOCs. (And, let me highlight it is an open question if they should be; my education science background is so weak that there are much smarter people to argue if this is good or not.) What dies this mean in terms of designing the service?

First we need to understand what kind of social interaction is needed. This understanding can emerge either by observing what are called lead users or by examining pedagogy and making an educated guess from there. The former means observing your users and checking on activates they already engage using different tools, like Facebook groups, IRC etc. where as the latter is more intervention to increase learning from the material by adding these social elements.

The next natural step is to develop this new idea but it is as important to validate the original idea: does the social element actually improve learning. Here is where tricks from the human-computer interaction may be handy: a small scale experiment may be enough to know if you’re moving into a helpful direction.