Skip to content

Group work advice for MOOC providers

The most valuable aspect of MOOCs is that the large number of learners enables the formation of sub-networks based on interested, geography, language, or some other attribute that draws individuals together. With 20 students in a class, limited options exist for forming sub-networks. When you have 5,000 students, new configurations are possible.

The “new pedagogical models” (A Silicon Valley term meaning: we didn’t read the literature and still don’t realize that these findings are two, three, or more decades old) being discovered by MOOC providers supports what most academics and experienced teachers know about learning: it’s a social, active, and participatory process.

The current MOOC providers have adopted a regressive pedagogy: small scale learning chunks reminiscent of the the heady days of cognitivism and military training. Ah, the 1960′s. What a great time to be a learner.

In order to move past this small chunk model of learning, MOOC providers will need to include problem based learning and group learning in their offerings. That won’t be easy. MOOCs have high dropout rates. Which means that if you’re assigned to a group of 10 learners, by the end of the course, you’ll be the only one left.

The large MOOCs can improve the quality of learning by creating a model for rapid creation/dissolution of groups. If you have teenagers in your house (or if you are a gamer), you’re likely familiar with how groups form in many video games or virtual worlds. There are two extreme opposites: World of Warcraft involves highly cohesive social units where individuals spend long periods of time together in solving problems and engaging in quests. In contrast, Call of Duty has low social cohesion as groups are formed on the spot and once a player logs off, the group dissolve (yes, you can log in and play with friends in a more cohesive unit on CoD as well). The latter model is worth considering for MOOCs.

Let’s say I take a course on Coursera. Due to high dropout rates, pre-planned groups will likely not work well. Instead, if I log in at 10 pm on a Friday in my statistics course, I can be automatically placed into a queue system similar to CoD: I wait until enough people show up to form a basic group, the system then launches us into our group work and we complete the 20-30 minute activity. If we like working together, we can decide to form a more stable group and schedule times to work online together. Otherwise, we disband. For the next group assignment, we are partnered with an entirely different group of learners.

To extend the group work experience, a quest layer can be added onto the assignment. Once a group is formed, each member is assigned a role that is vital to achieving a particular challenge. If members of the group don’t work together and share knowledge and skills, the problem will not be solved. The quest format will likely run longer than 20-30 minutes and may be most successful for groups that have worked together in the past.

My main point here is to emphasize that we need to think differently about group formation in learning when our learners have very weak social ties and when the commitment of learners to varies during the course. Taking a rapid group formation approach, augmented with quests, will help to ensure that some level of social learning occurs throughout the course, even after 90% of the learners have left.


  1. and we need to rethink the focus on a course, in contrast to a more community- or co-op-centric model…

    Sunday, March 10, 2013 at 10:39 am | Permalink
  2. dave cormier wrote:

    I’ve heard alot of people say that… but i’m not sure why they do. I love communities… and i love the ways they work… but they are extraordinarily hard to create and maintain.

    Community/co-op will only work in cases where people are near 100% passionate about what they’re learning. The vast majority of learning that happens is not in that category. I should totally learn how to do more SNA… but frankly i’m not passionate enough about it to participate in a community/co-op around it. I’m interested in the outcomes of SNA and the ways in which data analysis is being abused by people who are manipulating the data for their own purposes… not in SNA for its own sake.

    Hence the use of course. what say you?

    Sunday, March 10, 2013 at 10:49 am | Permalink
  3. Hi George,

    I follow your work with interest and you have been an inspiration for me since I discovered your book Knowing Knowledge when it was first published.

    I find your thoughts around group formation in open courses fascinating and I would agree with you that we need to re-think how this can be enabled in an open, rapidly changing and dynamic learning environment.

    I am currently experimenting with using Problem-Based Learning in open courses and organise an alternative open course, if you like, where Lars Uhlin a colleague educational developer from Sweden, and I, trial our COOL FISh framework which is in its infancy.

    If you would like to find out more about the project and the learning design, feel free to have a look at > We would love your comments. Thank you.

    Speak again soon.


    Sunday, March 10, 2013 at 11:19 am | Permalink
  4. so… focus on course because it’s hard to care about extra-course community contexts?


    Sunday, March 10, 2013 at 12:05 pm | Permalink
  5. dave cormier wrote:

    Focus on courses when you are trying to do courses and focus on communities when you’re working on communities.

    I don’t see them as mutually exclusive. I’ve worked very hard on community building over the years… and I also work on course building.

    Sunday, March 10, 2013 at 1:32 pm | Permalink
  6. Hi George,

    I too have been following your work since Knowing Knowledge was published. Thank you for your generosity and engaging so actively in the conversation.

    “we need to think differently about group formation in learning when our learners have very weak social ties and when the commitment of learners to varies during the course” Strengthening social ties is something I have been pondering – particularly how to improve the socialisation process when there are so many learners. “Introduce yourself” forums become overwhelming, and even the use of tags results in very large sub-groups.

    I think there are some lessons we can learn from online dating – I fill in a profile and the system recommends to me people I may want to meet for a virtual coffee. If I can make stronger connections, I am more likely to stay with the course and the group. I expand a little on it here:


    Sunday, March 10, 2013 at 5:39 pm | Permalink
  7. Jason Mock wrote:

    I like your idea, George, especially in how it flips the problem with forming groups on the head of how groups are traditionally formed in university courses. I’m thinking now that the ideal solution is one in which short-term groups can be formed and execute their task, but can potentially lead into longer-lasting groups. For those people who wish to invest themselves in this way, these longer-lasting groups have tremendous potential to positively influence their learning.

    To Dave’s point about building courses vs. building communities, I agree those are mostly separate. But I also think that, at least in a MOOC context, a goal should be to allow those who start off seeking a course but over time get invested enough that they wish to pursue a community, there’s a natural pathway available to them. Something as simple as attaching a Facebook group to the MOOC is a start, but I’m thinking a short-term group –> long-term group –> social community progression would work beautifully.

    But also the reverse: if someone dives into a community and realizes that they are perhaps not as well versed on the skills and knowledge shared by the community, an ideal situation would give them a natural pathway toward more of a course experience. Perhaps a progression like: starting in the large community –> finding a sub-community closer to your skill level –> discovering resources that target the level of your sub-community & discuss them as a sub-community.

    Exciting times!

    Monday, March 11, 2013 at 10:56 am | Permalink
  8. The social engagement aspect of learning has always made group learning more productive and efficient. I just sincerely wish that more educators around the world would really adopt the concept.

    The 1960′s was a great time for education! Educators really made students work hard during those times. I feel that even the best teachers now don’t really test students to there maximum level!

    Monday, March 11, 2013 at 11:06 am | Permalink
  9. I guess what I’m struggling with is lifelong-learning and student/learner-centric vs. institution-centric course-focus.

    If a learner is the centre of things, and if they are functioning in a time scale that’s longer than a single semester, how does the course fit into that? or the institution? (or a single institution)?

    I realize I’m overthinking things and following extrapolations that may be meaningless. But it feels like there’s something there, there…

    Monday, March 11, 2013 at 11:20 am | Permalink
  10. dave cormier wrote:

    @darcy The lifelong learning is the responsibility of the learning, i think, and not the person organizing course-style learning opportunities. Ideally, that learner would then take what they learned in a course back to their own community… but i don’t think we can build people’s communities for them.

    Monday, March 11, 2013 at 11:32 am | Permalink
  11. dave cormier wrote:

    uh… “responsibility of the LEARNER” not learning… doh.

    Monday, March 11, 2013 at 11:33 am | Permalink
  12. agreed. can’t build anyone’s community for them. just thinking about what it might look like for the pendulum to swing more to the individual side rather than institution. maybe it would be harder make sense of things, or to contextualize activities in a portfolio or program. I’m kind of hung up on the idea of a semester long course as the focus, rather than an ongoing flow.

    Monday, March 11, 2013 at 11:36 am | Permalink
  13. Coach Carole wrote:

    My experiences have shown me the value of collaborative learning in a MOOC. The model of encouraging and endorsing the natural formation of groups (small communities) will be part of our approach to the participation in our designing elearning MOOC. Indeed it has begun with the serendipitous grouping of designers who have now formed into a core group of developers. We are exploring early opportunities for participant groupings in networks such as LinkedIn and Google+. We are inviting the learners to look over our shoulders as we develop the course model, interactions and communications in a Mahoodle environment.
    Some advice please from this group: how would you encourage and structure collaborative group learning opportunities before, during and after the MOOC for an anticipated audience of educators (wanting to learn and/or fine tune their instructional design skills for elearning) around 300-500 in number?

    Wednesday, March 13, 2013 at 8:10 pm | Permalink
  14. I love the idea of a system that could create ad hoc groups. I would love to see MOOC providers support more synchronous technology in general. It would be great to have the possibility of easily having a spontaneous study group without pulling together a Google hangout or similar over a discussion forum!

    Thursday, March 14, 2013 at 10:48 am | Permalink
  15. Dmitry wrote:

    Interesting discussion. I am writing a blog-post about my experience with MOOC and e-Learning as a whole.
    Here is some controversy for me. I was really enjoying group-work during my study at the university and I miss in now. But from the other side I have 5-month old son, so I may study “when he allows”, not “when I would like” in the evenings and weekends.
    Anyway, hope to be able to plan my learning-time in the future.
    MOOC providers have statistics about everybody’s achievements like “completion rate” and “involvement level”. It is possible to build groups out of people who already proved their commitment to learn. “Statistics exchange” between MOOC providers would give advantages to such a group-formation principles.
    I have studied and am studying using Coursera, Udacity and Udemy. Will add more platforms to my experience in the future.

    Tuesday, March 19, 2013 at 12:43 am | Permalink
  16. Rolin Moe wrote:

    I know there is a ton of research on social learning as seen in WoW; I haven’t seen the same level in CoD. If you are using Wenger/Bandura/Vygotsky as a foundation, does a one-off group produce the community necessary for social learning to emerge, or is it another step toward personalized competencies and utilizing peers as nothing more than resources?

    Tuesday, March 19, 2013 at 12:16 pm | Permalink
  17. Rolin Moe wrote:

    BTW, above comment is only critical in the way research should question ideas. Truth be told, I think the idea is phenomenal and worth a hella trial run.

    Thursday, March 21, 2013 at 10:09 am | Permalink