A Human-AI Collaborative Precision Tutoring Model for Large Blended General Education Courses

Abstract

: General education arts courses in universities often face challenges such as large class sizes, a lack of precision tutoring mechanisms, and unmet demands for personalized learning. Drawing on Human-AI collaboration theory and precision teaching theory, this study constructs a three-stage, four-step precision tutoring model characterized by “data-driven — Human-AI collaboration — iterative optimization.” Taking the online elective course Fundamentals of Photography as a practical context, the study integrates a classroom behavior analysis system and an AI-powered artwork analysis tool into instruction. Over one semester, data on online learning behaviors, classroom behaviors, and AI-based artwork analysis were collected across four offline tutorial sessions to evaluate the model’s effectiveness. The findings reveal that the precision tutoring model moderately enhances student classroom engagement and the quality of their photographic work, with significant improvements observed in technical dimensions (composition and exposure) but not in artistic dimensions (color and thematic expression). Students acknowledged the division of labor in the model, in which AI handles automated analysis of technical dimensions while the teacher focuses on guiding creative and aesthetic aspects. The introduction of technology also led to a triple transformation in the teacher’s role: from “knowledge transmitter” to “data analyst,” from “unified lecturer” to “differentiated instructor,” and from “technology user” to “Human-AI collaboration designer.” This study offers practical insights for the intelligent transformation of large-class blended teaching in universities.