Economic stress, social isolation, and uneven experiences with remote learning are all contributing to some troubling trends in K–12 and postsecondary institutions, including decreased academic performance and sharp disenrollments. These trends point to a clear need for increased academic support to help students persist in school and progress in their learning.
Late last year, I explored two categories of online peer-to-peer learning tools that could help meet the need for academic support. One set of tools for peer teaching connects students who have mastered content and skills with those needing help. Another set of tools for peer learning has fellow students collaborating with each other for reciprocal learning benefits. In both cases, students learn from each other, but the difference is whether the tool distinguishes those giving help from those needing help.
Skeptics might ask: is peer-to-peer education really such an encouraging prospect? The idea of students teaching students might sound, at worst, like the blind leading the blind.
But there are reasons to be optimistic about peers as sources of academic support. Students are far more plentiful than teachers and professors, so there’s an opportunity to use their capacity to help meet the need for academic support. Additionally, peer learners’ commonalities can make it easier for them to intervene and support each other’s learning compared to teaching faculty. For example, some students might prefer to get help from someone who has recently taken the same class, or who shares elements of their identity or experience.
Nevertheless, optimism doesn’t always reflect hard reality, and peer-to-peer learning tools have to demonstrate that students are actually getting the support they need in order to make headway in the market. Fortunately, developers are pursuing a variety of strategies to improve the performance of these tools. Here are some of the different strategies that developers of peer teaching tools and peer learning tools are using toward improvement:
Tools for peer teaching: Solving for a technical problem
For peer teaching tools that match students with mastery to those needing help, the task of improving a tool’s performance hinges on a) facilitating matching, and b) vetting expertise.
Facilitating matching turns out to be fairly easy when content areas are clearly defined. For example, Knack, a peer tutoring platform for colleges, promises to match students needing help in a particular course with tutors who took that exact course. Brainly’s homework help tool allows students to search by question or browse by subject to either ask or answer questions.
Vetting expertise is a more formidable challenge, and companies are experimenting widely. Knack tutors need to have achieved good grades in the specific course they tutor for. StudySoup, where students exchange course notes, requires users to apply to become “Elite Notetakers” before earning money for sharing their notes. Vygo, a platform connecting college students to support services including tutoring, makes use of user ratings and reviews to help users sort through options. Brainly has human moderators to watch for bad answers to homework questions (although a glance through the site’s library suggests there’s a high volume of low quality responses that survive).
Some companies are also trying to solve the problem of guaranteeing peer-teacher expertise by not only vetting expertise, but building it. Knack offers an extensive library of training modules on tutoring best practices. Vygo is also beginning to develop training materials in response to customer demand.
Tools for peer learning: Solving for a pedagogical problem
Because peer learning tools make no distinction between students who are learning and students who possess expertise, something—or someone—must help spark students’ reciprocal learning. Hence the strategies companies are using to improve quality look quite different.
MIT professor Justin Reich has argued that online peer-driven learning, such as in communities like Scratch for block-based coding, can lead to extraordinarily deep learning but with uneven reach. Some students link up with peers and embark on projects that result in their developing substantial new skills, while others never find that spark. This means most classroom implementations will require a facilitator to water the seeds of inquiry between students.
Take, for example, the online debate platform Kialo, and social annotation tool Hypothes.is. Neither have features built-in to optimize for high-quality peer interactions. As a result, these peer learning tools can live or die by the pedagogical expertise of a teacher or facilitator. And unfortunately, most conventional teacher training programs focus far more on delivering instruction than on facilitating reciprocal learning.
One way developers are combatting this problem is by helping build teachers’ skills to use their tools in combination with effective pedagogy. Hypothes.is, for example, promises to help educators in its pilot programs develop teaching strategies that encourage collaborative and active learning—an exception to the norm, when most edtech companies restrict their teacher training to the tech itself.
Taking a different tack, other companies are developing technology to mitigate the amount of teacher effort that peer learning tools require. Packback, an online discussion forum tool, uses artificial intelligence to prompt students to ask deeper and more open-ended questions, reducing the burden on professors to stimulate discussion. Peerceptiv, a peer feedback tool, uses an algorithm to maximize the aggregate quality of all the feedback students receive, ensuring that everyone receives some robust feedback.
I’ll be keeping tabs not only on how peer teaching tools and peer learning tools are improving, but on the circumstances in which they’re being implemented. When large numbers of people can’t access a good or service, it’s an opportunity for innovators to scale access to more affordable and simple solutions. In a moment where academic support is not only scarce but unevenly distributed, peer-to-peer learning is a phenomenon to watch.