How to develop effective feedback within the teacher education programmes: comparing peers, experienced teachers and generative AI

Abstract

Generative artificial intelligence (GAI) systems are offering new opportunities to develop formative assessment activities that provide several forms of effective feedback. There are many chances and challenges linked to the use of generative tools. Teacher education programmes represent a crucial context where student teachers can experiment, analyse and investigate strengths and weaknesses. This study involved 86 pre-service teachers enrolled in the second year of the teacher education programmes at the university of XXX within the subject named «Curriculum development» during the a.y. 2025/26. Participants designed a series of activities for primary school and each student received three kinds of feedback: from the peers, from a tool of GAI, from an experienced teacher. Participants filled in a semi-structured questionnaire for each feedback. The instrument was aimed at investigating the perception of efficacy for each feedback. Especially, the participants highlighted whether and in what ways each feedback presents the features of a good feedback, sustaining self-regulated learning. Findings indicate that the three types of feedback contribute differently to the improvement of instructional design tasks. Peer and experienced teacher feedback appears more effective in supporting professional growth, whereas GAI offers detailed analyses and useful suggestions, despite sometimes it generates comments misaligned with the assignment. The most critical dimension is represented how feedback can support learning processes. In this way, it is necessary to educate future teachers to set up well structured feedback practices.