So the task is to design a MOOC that leverages the best of both worlds – xMOOCs and cMOOCs. George Siemens put together a team to look at this possibility for the next MOOC he is designing, and had a meeting called “Design Jam.” Since he works at my University now, I was able to beg/plead/bribe my way on the team. The biggest thing I learned from the Design Jam?
But aside from that, we have a lot of work ahead of us. The main design issue seems to revolve around having multiple paths through the content, mostly focused on creating a connectivist, learner-centered group work approach for those that prefer it, and also an instructor-centered path that guides the learners through the process for those that want that.
So the basic idea is that learners would enter the course and be presented with the option of going through one of the two routes. Maybe at some point an Artificial Intelligence data-driven program will even be able to recommend the path for them. Learners would enter one of the two paths and follow the paradigm presented. At any time that the learners on the cMOOC track need help (or at some point, when the AI data identifies a need), they can be directed towards the appropriate part of the xMOOC track for help. At any time the learners on the xMOOC track start to get comfortable with the idea of interacting with others (or the AI data identifies this), they can move into the cMOOC track. These movements could be a one time switch at any point, or a constant movement back and forth depending on the learner. Or the leaner could stick on the track they prefer the most. Or do both. Or lurk on one or the other or both. The system would basically look something like this:
(edit – not sure why I designed the original image from right to left. But click on the image for a larger version)
The idea is pretty straight forward, at least at a conceptual layer. This is an idea that I have been batting around in my head for a while and that many others at the Design Jam identified.
The technology behind it is another question.
The xMOOC path is pretty much in place with EdX. They have a good module-based system for presenting and assessing instructivist knowledge. Add on top of that they have connected to other systems through single sign-on and they are down with APIs… they have a system that is ready to connect with other systems as well as allow learners to move in and out as need with ease.
The cMOOC system that sits alongside that? That is another beast. Technology exists to create a learner-centered system (see A Domain of One’s Own)…. but how does this scale to possibly tens of thousands of learners?
Dave Cormier spoke of a system of community managers that he has found success with and that reminded me of something I read about the largest church in the world several years ago. This church in South Korea has close to a million members, yet connects every one of them to the community through a system of small groups that they call cell groups.
The idea of cell groups is an interesting one because it is based on the idea of organically formed groups that change, grow, die, combine, and otherwise fluctuate as needed. They can form based on location, shared interests, existing relationships, common goals, etc. The groups basically process the teachings of the church together and what they mean for their lives. If people join the church, they can join an existing cell group or form a new one. If existing ones grow too big, they can multiple into two or more. If a group dwindles, it can shut down and the remaining members can join other ones. Each group has about 8-20 members and one or more volunteer leaders that guide the group and run weekly meetings.
Every 8-20 groups are organized into sections, with the volunteer cell group leaders in the section meeting once month to go over issues. Every 8-20 sections are group together into districts, and so on. After a while this may not apply to education in a course that runs for maybe 5-8 weeks. But the idea would be to create a support system for the volunteer cell group leaders that could be based on, say, Teacher Assistants (aka “section” or “district” leaders) instead.
So the idea could be to organically form cell groups in the cMOOC, with each group forming based on location, shared interests, existing relationships, common goals, or whatever they like and they self-selecting a group leader. Roles for group leaders can be laid out a head of time. These could then be further grouped into sections under TAs as needed to deal with bigger issues that may arise. groups can then work together, grow, multiple, dies off, etc as needed for the life of the course, sending issues to the TAs as needed. Volunteer cell group leaders could meet in groups occasionally or as needed for guidance and help.
The problem comes with the software needed to do this. Kin Lane was brought to Design Jam to discuss APIs, and the idea of learner profiles was brought to the table during this discussion. These profiles could be used to help learners identify interests, goals, relationships, etc. Learners could then use their profiles to start forming groups as well as identifying these groups to the profiling system. The data behind the profile system could also identify potential group members. As groups grow, they could multiply (especially as new learners enter from the xMOOC strain). As other groups dwindle from drop-outs, existing members could use the system to identify new groups to join.
This software would also need to identify and map the cell group system in order to group cell groups into sections. It could also identify outliers that haven’t joined a group and see if there is an issue (some may just want to lurk, but others could be confused).
Further design of this system could even create a system for creating interactive spaces that don’t rely on third party products like Google Hangouts or Skype. Not that those are bad, but a lot of important data could be lost in those systems. If something like WebRTC could be integrated into this API driven profile system, learners could form, interact, and leave groups as needed throughout the lifetime of the course and just use the profile system to interact with video, text, etc through their browsers instead of a third party service. Since the cMOOC, xMOOC, social, and pedagogical systems would all be connected, massive amounts of helpful data could be collected throughout the entire class, further refining the system down the road. What leads to new group formation? What leads to groups dissolving? What leads learners to switch from cMOOC to xMOOC or vice versa? On and on.
This is not a fully realized idea or system. But its like we are working on the Sharknado perfect killing machine combination of xMOOCs and cMOOCs. Interesting stuff.