Unpacking Nursing Students’ Metacognitive Strategies in GenAI-Assisted Peer Feedback: Insights from Epistemic Network Analysis

This study used Epistemic Network Analysis (ENA) to examine how nursing students used metacognitive strategies during GenAI-assisted peer feedback. Sixteen undergraduate nursing students joined a structured feedback task. Each student used a Moodle platform supported by generative AI that helped guide planning, monitoring, and evaluating while giving feedback. Based on expert ratings, students were divided into high- and low-quality feedback groups using a median-split method. The ENA results showed clear structural differences between the two groups. Students who produced high-quality feedback built an active and connected metacognitive cycle that linked “Validating (VD)”, “Executing Strategy (ES)”, “Review (RW)”, and “Revision (RN). Their work reflected a habit of checking, reflecting, and adjusting ideas while using GenAI support. Students who produced low-quality feedback tended to stay in the planning dimension, including “Setting Goal (SG)” and “Making Plan (MP)”), and rarely engaged in monitoring or evaluating. This pattern suggested a form of cognitive outsourcing, where students relied too heavily on AI to guide their responses instead of reviewing their own reasoning. The findings show that metacognitive strategies can shape how students engage with GenAI tools. The study underscores the importance of designing pedagogical interventions that integrate metacognitive scaffolding to sustain learners’ agency, foster reflective feedback practices, and ensure that GenAI serves as a catalyst rather than a substitute for deep clinical reasoning.