Social-learning engineering

The social component of learning has long been overlooked from both a regulatory and a design perspective, with community formation often assumed to happen through the traditions of brick-and-mortar institutions. But as students spend less time at physical campuses, whether due to part-time status, family and work commitments, or online classes, deliberately planning how students will connect meaningfully with each other becomes necessary.

Coursera’s partnership to create “learning hubs” offers one example of how the education, business, and government worlds are exploring solutions to strengthen the tenuous social fabric that keeps students in class. Along with the basics of internet and technology access, these hubs also offer a more fundamental reason to return: social ties. Fellow classmates can offer instrumental support by sharing knowledge and experiences, but they also offer emotional support and validation when uncertainty strikes. While the time and effort required to build social ties may initially seem costly, the investment can pay off through higher enrollment and retention, as well as improved learning and satisfaction.

As these initiatives reveal, personalizing learning effectively goes beyond mere individualization to include genuine integration of the participants as people connected in a community.

MOOCs plus big data

In The Coming Big Data Education Revolution, Doug Guthrie argues that “big data”, rather than MOOCs, represent the true revolution in education:

MOOCs are not a transformative innovation that will forever remake academia. That honor belongs to a more disruptive and far-reaching innovation – “big data.” A catchall phrase that refers to the vast numbers of data sets that are collected daily, big data promises to revolutionize online learning and, in doing so, higher education.

I agree that there are exciting new discoveries and innovations still yet to be made through the advent of big data in education, and I also agree that MOOCs’ current reliance on scaling up delivery of existing content isn’t particularly revolutionary. Yet I see the two movements as overlapping and complementary, rather than as competing forces.

While MOOCs may not (yet) have revolutionized instruction, they have revolutionized access for many learners. Part of their appeal for those interested in their growth is their potential for enabling large-scale analysis due to the high enrollments as well as the availability of online data. The opportunity to study such large numbers of students across such disparate contexts is rare in traditional academic settings, and it permits discoveries of learning trajectories and error patterns that might otherwise get missed as noise amidst smaller samples.

Another potential innovation which traditional MOOCs (xMOOCs) have not yet explored is new models of building cohorts and communities from amidst a large pool of learners, a goal at the heart of “connectivist MOOCs” (cMOOCs) that highlights peer-learning pedagogy. Combine xMOOCs and cMOOCs, and you can improve educational access even further by enabling courses to spring up whenever and wherever enough people, interest, and resources converge. Add in the analytical power of big data, and then you have the capacity to truly personalize learning, by providing both the experiences that best support students’ learning and the human interactions that will enrich those experiences.

Alternate models for structuring learning interactions

Timothy Chester ponders the power of many-to-many peer networks in facilitating learning:

If there is to be a peer-based, many-to-many collaborative structure ensuring rigor and the mastery of learning outcomes, it must also be deemed authoritative and persuasive by participants. Some ways to ensure authority and persuasiveness might include the following:

1. The teacher must drive the collaboration. While teachers engaged in many-to-many relationships with students are not the authoritative center of the collaboration, they are responsible for structuring the student experience and stewarding the learning processes that occur.
2. The collaboration has to be bounded by a mutually agreed upon scope and charter. Compared to traditional one-to-many collaborations, many-to-many forms can appear chaotic or disorganized. In order to drive effective learning, many-to-many collaborations must operate within a set of boundaries – those things we might define as learning objectives, outcomes, standards, or rubrics. As steward of the learning process, the teacher must take responsibility for structuring the learning collaboration within a set of consistent and firm boundaries that include these structures.
3. There must be incentives for full student participation. Critics of peer grading systems in MOOCs note that such interactions by students many times lack significant investment of time and focus – resulting in peer feedback that is spurious. Both the quality and the quantity of peer feedback within a many-to-many system have to be statistically significant in order to avoid such spuriousness.

There are many models of such networks in both formal and informal learning settings: peer review systems (e.g., Calibrated Peer Review, SWoRD peer review, Expertiza), tutoring and peer learning communities (e.g., Grockit, P2PU, Khan Academy, OpenStudy), Q&A / discussion boards (e.g., StackOverflow), online communities (e.g., DIY, Ravelry), and wikis. The challenge for formal learning environments is to foster and nurture the kind of authentic, meaningful social interactions that emerge from sustained interaction within informal communities, in the context of the top-down and often short-lived peer experiences typically associated with school classes. Yet for personalized learning to succeed on a large scale, it needs to solve this problem effectively, so that learners are not isolated but can benefit from each other’s presence, support, errors, and wisdom.