How to scale peer learning online

Helping people learn with each other is the most important thing for P2PU to get right. In a few weeks we will be launching “Webmaking 101” to try new ways of helping (many) more people learn together. It’s a pilot project for the Mozilla School of Webcraft at P2PU that is based around the concept of “learning challenges” and it has potential for other communities within P2PU. Learning challenges mix great content with social support and mentorship to create a P2P learning model that can scale in ways that traditional courses can’t.

This is a long post, because I felt it was important to talk a bit about our experience so far, and how the idea of “challenges” is a direct response to some of the things we have learned. I hope the post provides a good starting point for a conversation, so that as a community we can explore how P2PU can best support peer learning in the future – through courses, challenges, and other models that we haven’t even thought of yet.

Why this is important

We know that peer learning works offline. At Harvard University (of all places) the ability to get help from other students through informal study circles is the most important predictor of academic success among undergraduates. We learn more when we share our questions with others, and try to explain things that we just learned to them. In this way it is not just content we learn, but we get better at the process of learning itself.

The Internet supercharges the potential of peer learning. It gives us access to a huge (and growing) amount of high quality educational content that can be downloaded, translated, remixed, adapted and shared. And it puts us in touch with millions of other people to learn with. It is this combination of peer learning, the social web, and educational content that creates an unprecedented opportunity to make education more accessible, more engaging, more fun, and less expensive.

This is why figuring out how to support peer learning online is important – and why it has shaped everything we have done at P2PU, and how we have gone about it.

Peer learning online is hard

When P2PU was just a bunch of idealists and a hosted wiki, we were overwhelmed by the positive response to our simple idea – let anyone create and run a course online. People were excited by the opportunity to share their knowledge, and many joined our community and created great courses. But as we grew we started noticing a few challenges that seemed related to structuring learning as courses:

  • Not all courses promoted peer-learning > Some of our courses turned out like more traditional instruction, with experts leading the discussion, answering most of the questions, and pointing out the right from the wrong. That’s great for some learners and topics, and we are not abandoning it, but it doesn’t scale like peer learning and it doesn’t encourage learners to take active ownership of their learning.
  • Drop-out rates were high -> In retrospect this should not have been surprising, because drop-out rates are high in all online learning (even in courses where students pay and are working towards a formal degree) but it’s frustrating for volunteer facilitators to see participants leave, and it’s discouraging for learners to loose their peers.
  • Courses didn’t run consistently -> We have had many great courses run once, or twice, but rarely more than that. And yet there was a seemingly infinite demand for some topics (I’m exaggerating, but it did make our server crash a few times) and there is nothing worse than not being able to meet that demand.


We realized that our model of courses placed a huge burden on the role of the facilitator, and that it is hard to facilitate peer learning online. It is even harder in an environment like P2PU where many of the usual incentives like fees and degrees don’t exist.

Great facilitators are hard to scale. One reason is the high level of commitment that is required. Many people are willing to help others learn, but fewer have the spare hours to create and offer an entire course. And offering the same course again, or offering a course that someone else designed, is less interesting than creating a new one.

And finally we realized that great content that is designed to support peer learning is scarce. When we started out we thought the content problem was solved. The reality is that there is a lot of content, but a lot of it isn’t engaging for self-learners, or easily suitable to support informal peer learning communities.

From courses to challenges

The realization that there may be a problem with the existing content led us to think about what we could add to the wealth of materials that exist already. Let me be clear, there is no shortage of open educational content, and much of it is very good. And we have no intention to produce more of what is already out there. However there are ways to make what’s out there more useful to peer learners, by turning content into learning challenges and:

  • Choosing the best, most relevant and interesting open resources
  • Framing them with the interesting questions that spark curiosity
  • Setting tasks that emphasize peer-learning and project-based-learning
  • Defining clear goals that signal achievement, with enough flexibility to take different paths to reach them
  • And ultimately involving the community into the development of more and better challenges add
  • Providing mechanisms for assessments and badges, and signal achievement
  • Weaving social learning features into and around the content

There is an important content foundation to the “challenge” model (also see this post by Chloe on what makes a great challenge), but the most exciting thing is not the content itself, but the social learning that it enables.

Peer learning that scales

Having great challenges makes great facilitation easier. It allows self-learners to get started on their own or in informal cohorts. It let’s us lower the bar for more people to get involved to answer questions, or act as mentors, which is much easier and takes less time than signing-up as a course facilitator. Learning cohorts can be organized around specific challenges, nudging everyone currently working on a particular challenge to support each other. And it’s straight-foward to attach badges to challenges, which act both as motivators to work harder, and map out a path along which to progress.

As a result many more people can get involved in helping others learn. And many more people can learn. 

This does not mean that we have plans to abandon the course model. Quite the contrary, we will continue to get better at supporting it – by providing better resources to new facilitators, rolling out improvements to our site, and thinking about ways to motivate facilitators to run their courses more than once. But it’s time for something a little more radical.

Taking it to the users

In true P2PU style we decided that the best way to find out how challenges would work was to go ahead and try it. John, Jamie, Erin, Chloe, Zuzel Jessica (and many others) have done a really awesome job building out the first set of challenges for Webmaking 101, which will launch in a few weeks. We asked the Webcraft community for feedback on earlier drafts, and they have had lots of useful input which helped us make the challenges better.

But we want to broaden the conversation beyond Webcraft. I am excited about the challenge model because I think it offers an opportunity for P2PU to scale great peer learning to many more people than we can reach with courses. But I also have many unanswered questions about challenges. Here are three for starters:

  • How can we preserve the strong sense of cohort that traditional courses have?
  • Will challenges only work for technical areas, or can they be applied to other fields as well?
  • We created the first set of challenges for Webmaking 101. How can more people get involved in creating more challenges?

Lots of hard questions. Let’s experiment, find answers and keep getting better at helping everyone learn with everyone else.

Cheerio – P


Back to Blog home