AI-Driven Feedback in Graduate Education: ChatGPT Offers Detailed Guidance but Lacks Personalization

• In Event: AI in Education

Abstract

This study explores the use of ChatGPT for providing feedback in a graduate-level asynchronous online course. Using the PERFECT model to design effective AI prompts, it investigates how AI can provide detailed feedback, act as a feedback guide, and offer improvement suggestions. The research gathered qualitative and quantitative data to understand students' perceptions, experiences, and satisfaction with AI feedback. Findings show that while AI feedback is actionable and beneficial for idea generation and guidance, it lacked personalization and contextuality. The study highlights the potential for integrating AI into teaching practices to enhance feedback processes and offers practical guidelines for educators. Future research should focus on refining AI prompt design and exploring AI-human feedback integration for improved educational outcomes.

Authors

• Luis E. Pérez Cortés, University of Texas - Arlington

• Hugh Kellam, University of Texas - Arlington

• Tranell Gilmore, University of Texas - Arlington