Peer feedback modelling needs in online higher education: An exploratory study

Peer feedback (PFB) is considered a reliable and beneficial method for supporting learning and interaction. However, PFB is a complex skill that needs explicit instruction through modelling. To leverage the positive effects, a theoretical and pedagogy-driven application of PFB - namely scaffolded PFB modelling - was integrated into an online higher education academic writing course. To address needs-driven modifications of PFB within this course, students were surveyed after course completion to provide insights into their additional supporting needs about PFB. In total, 173 students answered reflective questions and six students were interviewed. Their responses were analyzed using Atlas.ti. Analyses showed that the need for PFB modelling was especially shown in the phases of observation and emulation. Students indicated the need for examples, additional (step-by-step) instructional materials, and the opportunity for guided practice and additional teacher feedback. In addition, students also provided suggestions to improve the PFB process; most of these suggestions were related to the online aspect of the course and its negative influence on interaction possibilities, which impeded communication and the development of trust and relationships among students. These results demonstrate that a pedagogy-driven application of PFB in an online higher education course requires a needs-driven, context-dependent redesign of PFB assignment(s), especially in the observation- and emulation modelling phases with extra attention to interaction. • (Online) PFB is a complex skill that requires explicit instruction, particularly through scaffolded modelling. • An online course can present challenges in interaction and trust-building, which are essential for effective PFB. • Students often struggle to give effective (written) feedback due to low confidence and unclear evaluation criteria. • Instructional support, especially in the observation and emulation phases of modelling PFB, is important.