Exploring a Metacognitive Scaffolding-Based GenAI-Assisted Peer Feedback Provision Approach to Enhance Feedback Engagement Among Nursing Students

Background: Providing effective peer feedback is a challenge in nursing education. While Generative AI (GenAI) can assist, students often struggle with the task. Metacognitive scaffolding may help guide students through this complex process. Aim: This study aimed to evaluate the effects of a metacognitive scaffolding-based GenAI-assisted peer feedback provision (MGPFP) approach on nursing students’ feedback engagement and behavioral patterns. Methods: A quasi-experimental study was conducted with 71 nursing students. The experimental group (n = 35) used the MGPFP approach, while the control group (n = 36) used a standard GenAI-assisted approach without scaffolding. A Mann–Whitney U test was used to compare feedback engagement. Lag sequential analysis was used to examine feedback giving behavior patterns based on coded video data. Results: The experimental group reported significantly higher engagement than the control group across four dimensions: behavioral, cognitive, social, and emotional engagement. The experimental group generated 5219 coded behaviors, while the control group generated 1861. In the experimental group, common behaviors included referring external resources (19.58%), comparing and making judgements (17.80%), and recognizing the purpose (15.77%). Non-feedback behaviors were much higher in the control group (2.69%). Lag sequential analysis identified 17 significant sequences in the experimental group and 14 in the control group. Conclusions: Integrating metacognitive scaffolding into GenAI-assisted peer feedback can improve nursing students’ engagement and promote more productive and structured feedback behaviors. This approach is a valuable strategy for enhancing the quality of peer feedback in nursing education.