Sequencing matters:Exploring AI and peer feedback in undergraduate physics presentation preparation

Research topic/aim This quasi-experimental study investigates how sequencing of AI and peer feedback influences students’ engagement, perception, and revision practices in an undergraduate physics course. The study aims to support students' preparation for a novel oral exam format, where they present a solution to a physics problem. It explores how different feedback sources are used depending on their placement in the learning sequence and the type of feedback they provide. Theoretical framework Approaching feedback as a co-constructed process (Winstone et al. 2022), student feedback literacy is often hailed as an important competence in higher education (Carless & Boud 2018). Students’ appreciation of feedback, ability to make judgments, manage emotions, and act on feedback are thought integral to learning with formative feedback. Thus, developments in artificial intelligence increases focus on AI’s potential as a feedback source, particularly in how it complements or differs from human feedback and relates to feedback literacy. Methodology/research design Twenty students were divided into two groups: one group received AI feedback prior to peer feedback, while the other had the order reversed. Each student brought a prepared solution to a specific problem, intending to revise and later present it in class. Students spent 10 minutes engaging with a chatbot prompted to give high-quality feedback on solutions and presentations, either before or after 10 minutes of peer feedback. Data sources included pre- and post-surveys, chatbot interaction logs, observations of peer feedback, and five interviews. Expected results/findings The sequencing of feedback shaped students' experiences and outcomes, as some students adapted their use of the second feedback source based on their experience with the first. Peer feedback, when provided first, appeared to frame subsequent AI feedback as more task-focused. Students used the relational and situated nature of peer interaction to focus more on presentation aspects. While AI feedback offered consistent, detailed commentary, some students experienced it as overwhelming or irrelevant. Nevertheless, other students reported that even a single sentence from the AI could trigger meaningful reflection and revision. Peer feedback quality varied considerably, ranging from lacking substance to being highly engaging. Nordic relevance This study explores an instructional design where AI supports the collaborative nature of classroom teaching in Nordic higher education, which traditionally emphasises collaborative learning and peer feedback. The study highlights the importance of sequencing in feedback design, showing that engagement with feedback is not isolated from prior experiences with other feedback sources. While AI feedback offers consistent scaffolding, peer feedback can provide contextual and affective support. Educators designing feedback interventions involving both human and AI sources should consider how sequencing influences students’ uptake and perception of feedback, especially in preparation for high-stakes assessments. References Carless, D., & Boud, D. (2018). The development of student feedback literacy: Enabling uptake of feedback. Assessment & Evaluation in Higher Education, 43(8), 1315–1325. https://doi.org/10.1080/02602938.2018.1463354 Winstone, N., Boud, D., Dawson, P., & Heron, M. (2022). From feedback-as-information to feedback-as-process: A linguistic analysis of the feedback literature. Assessment & Evaluation in Higher Education, 47(2), 213–230. https://doi.org/10.1080/02602938.2021.1902467