Open courses taking advantage of the affordances of new network technologies to enable people to learn together in a peer-to-peer fashion have been developing for several years, and are having an impact across the formal and informal sectors. An early experiment was Professor David Wiley, who opened his university class to the world in 2007, and spawned a number of similar experiments (Fini et al., 2009Fini, A., Formiconi, A., Giorni, A., Pirruccello, N. S., Spadavecchia, E., & Zibordi, E. (2009). IntroOpenEd 2007: an experience on Open Education by a virtual community of teachers. ; Couros, 2010Couros, A. (2010). Developing personal learning networks for open and social learning. Emerging Technologies in Distance Education, 109–128. ).

Inspired by this, the Canadian researchers George Siemens and Stephen Downes created the concept of Massive Open Online Courses (MOOCs), which has since spawned a number of connectivist MOOCs, and several research papers (for example Mackness, Mak & Williams, 2010Mackness, J., Mak, S. F. J., & Williams, R. (2010). The ideals and reality of participating in a MOOC. In Seventh International Conference on Networked Learning, Aalborg, Denmark. , Kop, Fournier & Mak, 2011Kop, R., FournierHélène, & Mak, S. F. J. (2011). A pedagogy of abundance or a pedagogy to support human beings? Participant support on massive open online courses. The International Review of Research in Open and Distance Learning, 12(7), 74--93.). Last year, Sebastian Thrun and Peter Norvig from Stanford launched a new generation MOOC in artificial intelligence, and attracted more than 150,000 learners. The two companies resulting from this experiment, Udacity and Coursera, as well as MIT and Harvard’s collaboration on edX, has made the concept of open courses and peer-learning known among a much larger audience Roscorla, 2012Roscorla, T. (2012). Massively Open Online Courses Are 'Here to Stay'. Converge. .

The Peer 2 Peer University (P2PU) is a community of volunteers that organize Open Educational Resources into courses, run by volunteers and open to everyone for free. Launched in 2009, P2PU has hosted hundreds of courses and thousands of learners in a large-scale experiment aiming to improve our understanding of peer-learning. There have been several studies of individual P2PU courses, and one study of the concepts of openness among users and administrators (see for example De Liddo & Alevizou 2010, Alevizou 2010, Corneli & Ponti 2010, Ford et al. 2011, Hofman 2010 and Corneli & Danoff 2011), but no comprehensive study of the diversity of courses offered at the platform, and their success rates.

Research method

The first round of P2PU pilot courses were run on two disconnected platforms. After students and course organizers complained that it was difficult to keep track of a course happening on two disconnected platforms, a course management system was developed on Drupal, which provided basic tools for each course: syllabus, sign-up questions and student management, sharing files and documents, a discussion forum, and a text chat (web IRC). This platform was available during one and a half year, and during that time 107 courses were offered for enrollment, and 50 courses were listed as drafts in progress. This study examined every course that was officially offered for registration during this period.

Of these, 21 never got off the ground, 7 migrated to a new platform in the spring of 2011, and 6 courses were primarily conducted on other platforms that are no longer available, so that no data could be extracted from them. This leaves 73 courses to be studied. All of these courses are available in archived form on http://old.p2pu.org. The researcher examined each course, looking for data on participation patterns by students, course longevity, instructor expertise and orientation, and tools used.

Courses examined

The courses spanned a great number of disciplines, although technology-related courses made out 40 of the courses (54%). There were also 7 courses on math (9.5%), 6 on social science topics (8.2%), 4 on copyright law (5.4%), as well as courses on foreign languages, creative writing, music, journalism, marketing, health, sustainable agriculture, education, psychology and cooking.

While most of the technology courses were introductory (Introduction to Python programming, Introduction to HTML5, etc), many of the other courses were targeted at a graduate level, and often had very specific titles, such as “Kitchen Science”, “Open Governance and Learning” or “Psychology of Math Learning”. Some of the courses were created by P2PU “Schools”. The three schools active during this period were the School of Webcraft, in collaboration with the Mozilla Foundation, which targeted teaching basic web skills to as many as possible, using open source tools. The School of the Mathematical Future aimed to share the joy of math teaching and learning, and the School of Social Innovation ran a number of innovative courses using local groups and external tools of communication.

Success rate

It is very difficult to define the success criteria for a P2PU course, since both course organizers and students might have very different goals in each course. There is neither a strong carrot, nor a stick, to keep students (or even the facilitator) in the course until the end - only the benefit of the social learning interaction. This means that P2PU courses have very large dropout rates, and many courses do not continue until the end.

Because all of the course materials and interactions are open to the world, it is also possible for students to study the materials by themselves, and follow the conversations in the class, without contributing anything themselves. The individuals can still learn from this process, but if too many people choose to do so, the course “dies out”, and there is nothing left to follow. (In the cases where the course organizer has assembled a complete curriculum for the course, individual students can choose to continue self-studying these materials - however, as they leave no trace, this kind of learning is invisible to the research methods used in this study).

Given these concerns, it is still interesting to consider how many courses “ran until the end”. In my study, I looked at whether there was constructive activity and participation from the course organizer and at least some of the students until the end of the course. I found that 35 courses (48%) ran until the end. Of the 38 courses that “petered out”, 11 (29%) never got off the ground, and died out within the first two weeks, 7 (18%) ran out of steam close to the end, and the rest petered out somewhere in the middle of the course. In five of the cases (13%), the organizer disappeared and the students actively asked about his or her disappearance. (In some courses, the students tried to organize themselves to continue learning without the organizer, which was much encouraged by the organization, and in some cases the organizer reappeared apologetic, yet the courses still floundered shortly afterwards).

It is important to take the numbers above with a grain of salt. Even the courses that did not run until the end were in some cases very active for at least a few weeks, and often received comments from participants who said they had learnt a lot. And courses that ran until the end often lost many of the students by the wayside, so that only a few of the original group were still active during the last week.

Role of the course organizer

P2PU course organizers are unpaid volunteers with diverse backgrounds. In many of the presentations about P2PU, the peer-to-peer learning aspect is underlined, as well as the idea that the “resources are the expert”. However, in practice almost all the course organizers in this study had expertise in what they were teaching, and most of the courses were strongly instructor-oriented. The degree of instruction varied – in some cases the course organizers produced large amounts of original instructional content, or used synchronous meetings to explain topics, in others they used external materials and functioned more as facilitators.

Among the courses, there were only 6 (8%) that can be categorized as truly peer-learning, without an expert course organizer who drove the course forwards. Some were based on existing external resources, such as an MIT OpenCourseware, or a textbook, others were more open-ended and asked students to suggest relevant resources and topics to discuss. Only one of these courses succeeded, whereas the others often started with great enthusiasm, but then petered out within a few weeks.

Constructivist Massive Open Online Courses are an interesting example of the challenge of purely “peer-learning”. They exist in a dynamic field between a charismatic expert organizer, and a larger number of distributed students with very little explicit guidance who are supposed to self-organize and learn collaboratively. Mackness, Mak & Williams, 2010Mackness, J., Mak, S. F. J., & Williams, R. (2010). The ideals and reality of participating in a MOOC. In Seventh International Conference on Networked Learning, Aalborg, Denmark. questioned the low number of participants (14%) in one course, and highlighted the tension between the principles of MOOC learning as autonomy, diversity, openness and connectedness, and the fact that the more these were present, the more likely participants’ learning was to be limited by lack of structure, support and participation.

As mentioned above, this does not mean that the courses were complete failures – some hosted quite interesting discussions before they petered out, and they often functioned as experiments to test different learning theories. For example, Joseph Corneli (2010) led two peer-focused courses, and based on his frustration with the lack of success, began developing a theory of how to learn in a peer-to-peer fashion, which he calls paragogy.

Given that P2PU is as much an experiment in social organization, as it is in learning theory, it will be important to continue experimenting with ways to foster true “peer-to-peer learning” that reduces the demand on the course organizers. The courses that were successful, were often so because of the unsustainable amounts of time that the unpaid organizers committed to them. In some cases, this led to burnout, and courses faltered when the course organizer withdrew in the middle, or it might have led to course organizers not volunteering to teach another course soon after having finished one. Course activity was largely supply-driven, in that every time a course was offered, no matter how esoteric (for example “Learning New Testament Greek, in Spanish”), students signed up, and there was a constant demand for many more classes than could be supplied.

Tools used

The survey of courses shows that every single course used the syllabus and discussion forum to some extent. Some courses worked mainly asynchronously, with discussions in the forum, whereas others also implemented regular synchronous meetings. Some courses used the text chat facility for this, whereas others used Skype, Elluminate, WizIQ or FlashMeeting. Some more familiar with other platforms used external wikis to host content, and several courses encouraged students to start a personal blog, and use this for reflections and projects. For collaborative writing, both Google Documents and Etherpad was used, and programming courses often used GitHub to share code.

A number of groups disliked the discussion forum, especially the lack of e-mail integration, and chose to use a mailing list hosted on Google Groups instead. One of the challenges of this study was that the content posted on external sites was in many cases not available anymore, individuals might have taken down their blogs, or entire Web 2.0 services might be unavailable.

The two most common models for organizing courses were either purely discussion based, where students would access some material each week, and then reflect on these and discuss, either in their own blogs, or in the forum. The other model was more task-oriented, and was very common in the more practically-oriented courses, where you were expected to follow a tutorial and solve some technical problem, which you could then share with the others. The copyright law courses used the case study methodology, and a few other courses also had assignments to groups of students. In these cases, the groups all found their own ways of collaborating, and might have used tools not listed above.

Conclusion

Going through all the courses, one is struck by the richness of material and interactions that have taken place. Many courses floundered, but one of the advantages of a platform such as P2PU is that it affords rapid iteration and risk taking that would be difficult to justify in a more formal setting. So far, most of the courses seem quite inspired by what students and organizers think of as “proper university courses”, with syllabi, tasks and in some cases, even grading. However, to be able to scale and truly fulfill the promise of providing free and quality learning opportunities to all, more research is needed on how to foster truly peer-learning without an expert, or with minimal involvement by experts and course organizers.

References