I'm in the midst of designing an online course using principles of peer-2-peer learning. Rather than set myself up in the usual role as instructor, the course is getting designed for the students to be the teachers. Obviously they cannot convey expertise, but they can benefit from the pattern for "the best way to learn something is to teach it". They will also benefit from playing both roles of the one helping and the one getting helped. This departure from the role fixity of formal instruction sets the learners up to help themselves more effectively as they pursue a path of further expertise beyond the course.
Rather than give students information, it's essential to provide tools to be used by them from the start. This sets up some initial hands on learning to become familiar with the use. It also generates some authentic interest in getting better informed once the context of use seems real. This approach complies with the pattern of "use it or lose it" by insisting that praxis precedes becoming better informed.
It's essential that each participant cultivate a unique contribution which leads to a distinguished reputation. Most academicians assume they only way to differentiate collaborative peers is by their expertise. They set up team projects in the classes without giving each student a unique role, resource to contribute or viewpoint to apply to the team group dynamics. This occurs because all the students uniformly lack academic expertise. This all-too-common approach to teamwork gives p2p approaches a bad name.
Peer contributions can be individualized by providing different resources to apply to common situations or the same resource to apply to differentiated use cases. The use of these resources gets greatly enhanced by peers making requests for help. A dialogue emerges about the context of the request which makes the resources seem increasingly useful, relevant and worthy of further practice. A meta level of thinking emerges to resolve questions of applicability, timing, balance with other issues, far beyond the conventional concerns with right answers and grades.
As peers make requests for help, the intrinsic rewards for formulating responses emerges. Requests set up the possibility of the responder receiving verbalized appreciation, useful feedback and deeper insights into how to be more helpful. None of that gets tainted by extrinsic rewards of grades or formal evaluations by the instructor. The hands on experience of deal with peer requests some self-generated criteria for what makes a better request and a more helpful response. There's no handout of evaluation rubrics to ensure consistency of evaluations. It's assumed the internal criteria will prove to be more enduring and applicable by each peer.
When these design principles get applied, the instructor can monitor all the activity, make occasional processes observations and learn from the interactions how to improve the design for p2p learning. The peers learn at more levels and in more useful ways to face future challenges.