Research on Behavioral Mechanism and System Design of Art Teachers Adopting AI Classroom Assessment Technology Based on Extended TAM
This study addresses the high subjectivity and heavy workload associated with traditional evaluation in art education, as well as the lack of attention to user adoption behavior in existing AI-in-education research. It aims to reveal the behavioral mechanisms underlying art teachers’ sustained use of AI-assisted classroom assessment technologies and to inform system design. An extended Technology Acceptance Model (TAM), incorporating information technology self-efficacy and AU (AI Application Intensity), was constructed and empirically tested using structural equation modeling (SEM) based on questionnaire data (N = 103). The sample primarily consists of art teachers with relatively high experience in using AI tools, and thus the findings are particularly applicable to active technology users. Results indicate that information technology self-efficacy serves as the initial driver of behavior and has the strongest effect on perceived ease of use (β= 0.59). Perceived usefulness significantly influences both BI (Behavioral Intention) (β= 0.39) and application intensity (β= 0.25). A chain mechanism from self-efficacy to ease of use, then to usefulness, intention, and ultimately application intensity is validated. Based on these findings, three design principles are derived, and a teacher-centered conceptual system architecture is proposed. The study provides a coherent linkage between user behavior mechanisms and system design, offering theoretical and practical implications for developing AI-based assessment systems that are usable, adoptable, and sustainable.