Abstract As Generative Artificial Intelligence (GenAI) increasingly mediates formative assessment, understanding how learners across achievement levels perceive the distinct roles of human instructors versus GenAI tools presents an emerging research gap. This study addresses this gap through secondary qualitative analysis of interview data from pre-service science teachers, stratified into high-, medium-, and low-achieving groups, who experienced ChatGPT-enhanced formative assessment in stoichiometry learning. The findings reveal distinct, achievement-based perceptions of instructor and GenAI roles. The human instructor was predominantly viewed as an adaptive expert, adept at toggling between the roles of a Simplifier and an Elaborator depending on the learners’ achievement levels. Conversely, ChatGPT was perceived as a personalized tool for self-regulated learning, with its role shifting according to student achievement: a Patient Tutor for low-achievers, a Personal Coach for medium-achievers, and an Intellectual Sparring Partner for high-achievers. Based on these findings, we propose an Instructor–AI Synergistic Learning Ecosystem model that reframes human–AI collaboration not as a competition but as a complementary partnership – transforming the long-standing challenge of scaling individualized formative assessment into a pedagogical opportunity for AI-integrated education.
Ratniyom et al. (Wed,) studied this question.