AI self-efficacy and student teachers’ engagement with generative AI feedback: Effects on reflection, feedback adoption, technostress, and learning performance

Although generative artificial intelligence (GenAI) has been applied in education, little is known about how student teachers engage with GenAI feedback in reflective learning, or how AI self-efficacy shapes this process. This study integrated GenAI feedback into digital storytelling for student teachers’ professional identity reflection. It examined the reflection level, feedback adoption, and self-evaluation of DST work of student teachers with high AI self-efficacy (HAISE) and low AI self-efficacy (LAISE). Group differences in GenAI literacy, technostress, continuous usage intention, and learning performance were also explored. The sample comprised 76 student teachers. Results disclosed that: (1) Most student teachers reached reflection level with the support of GenAI feedback; (2) The HAISE group tended to adopt GenAI feedback to improve their digital stories; and (3) The HAISE group outperformed the LAISE group in GenAI literacy, technostress, and learning performance. The findings have implications for AI self-efficacy, feedback engagement, and GenAI-facilitated reflective learning.