FEEDBACK IN THE ERA OF GENERATIVE AI

Abstract

Purpose: The purpose of this study is to investigate the perception of students and educators in higher education towards feedback provided by large-language model AI, with feedback and interaction with human educators. Theory: Drawing on the dialogism framework and considering the emotional and relational aspects of feedback, this study examines the role of dialogue and interaction in feedback processes and the implications of integrating AI-generated feedback with human feedback. Method: Adopting a qualitative research methodology, the study encompassed focus group interviews with 17 university students and nine educators across two esteemed Swedish universities. The gathered data was dissected via thematic analysis, spotting themes that resonate with the participants' experiences and views on feedback with generative AI and human instances. Results: The analysis revealed three main themes, highlighting various aspects of feedback in the context of using generative AI tools such as ChatGPT alongside human educators. These themes emphasised the importance of understanding the nature of generative AI and human feedback, addressing the emotional dimensions of feedback, recognising potential risks and ethical concerns of using generative feedback, and exploring the integration of AI-generated feedback with human feedback practices to enhance learning engagement and outcomes. The findings contribute to the understanding of the potential and risks of AI-generated feedback in higher education and inform the development of best practices for integrating AI and human feedback ethically.