Can Generative AI be used to deliver personalised yet sustainable feedback to large cohorts?
Lead institution: Keele University School of Medicine
Country: United Kingdom
Project primary contact: Nazim Ali, Lecturer and Deputy Director of Education (n.ali1@keele.ac.uk)
Project secondary contact: Sarah Aynsley
Project Title: Can Generative AI be used to deliver personalised yet sustainable feedback to large cohorts?
The project is: In Progress
Project Summary:
It is recognised that providing written feedback which is detailed and specific is time-consuming and requires it to be delivered in a timely manner for student uptake. However, providing personalised feedback is unsustainable especially within a large cohort of students. Thus, there is a need to find ways to provide personalised feedback. We are exploring the use generative AI (GIA) models as feedback tools to provide constructive and teacher-like personalised feedback. As proof of concept, we have developed a ChatGPT based system which can provide feedback on students’ work. We used a panel of teachers to evaluate our approach and they reported that the ChatGPT derived feedback was detailed and constructive. We perceive that a GIA based feedback system could offer significant pedagogical benefits such as real-time feedback which would maximise student engagement with feedback. As such we are currently undertaking a full-scale study to test our approach within a large cohort of students with the aim of gathering their experiences and perceptions of using GIA as a personalised learning tool.
Impact: To improve student learning by providing personalised and real-time feedback.
Audience: The project will directly impact students and how they engage with feedback.
This output is part of a member project - AI Garage: Creating the Future Now which collects and curates cutting-edge practice examples of AI. You can explore other submissions here.