Artificial Intelligence as a Catalyst for Educational Reform in Economics: A Structural Equation Modeling Approach to Understanding Student Adoption Behavior, Engagement Mechanisms, and Perceived Learning Gains.

https://doi.org/10.5281/zenodo.18174174

Abstract

The rapid integration of artificial intelligence (AI) into educational environments has generated significant interest in its potential to catalyze pedagogical reform, particularly in disciplines such as economics that require the synthesis of abstract theory, quantitative reasoning, and real-world interpretation. Despite the growing availability of AI-enabled learning tools such as conversational agents, adaptive tutoring systems, and intelligent feedback platforms empirical understanding of the mechanisms through which students adopt these technologies and translate usage into meaningful learning outcomes remains limited. This study addresses this gap by proposing and empirically validating a comprehensive structural equation model that explains AI adoption behavior among economics students and examines the engagement pathways through which AI use contributes to perceived learning gains. Drawing on established technology adoption theories and student engagement literature, the proposed framework integrates key antecedents of AI adoption, including perceived usefulness, ease of use, social influence, facilitating conditions, trust in AI, perceived risk, and AI self-efficacy. The model further conceptualizes student engagement as a multidimensional construct encompassing cognitive, behavioral, and agentic engagement, positioning engagement as a mediating mechanism between AI adoption and learning outcomes. Data were collected through a structured survey administered to undergraduate and postgraduate economics students across multiple higher education institutions. The measurement and structural models were analyzed using structural equation modeling, with rigorous assessments of reliability, convergent validity, discriminant validity, and model fit. The empirical results demonstrate that perceived usefulness, trust in AI, and AI self-efficacy exert strong positive effects on students’ intention to adopt AI tools for economics learning, while perceived risk negatively influences adoption behavior. Adoption intention is found to significantly enhance student engagement, particularly in terms of cognitive and agentic engagement, reflecting deeper conceptual processing and proactive learning behaviors facilitated by AI interactions. Engagement, in turn, exhibits a substantial positive effect on perceived learning gains, including improved conceptual understanding, problem-solving confidence, and perceived academic performance. Mediation analysis confirms that student engagement plays a critical intermediary role in translating AI adoption into learning benefits. By elucidating the structural relationships among AI adoption drivers, engagement mechanisms, and learning outcomes, this study contributes to the emerging literature on AI-enabled education and provides actionable insights for educators, curriculum designers, and policymakers. The findings underscore that effective AI-driven educational reform in economics depends not merely on technological access but on fostering trust, competence, and engagement-centered learning practices that enable students to meaningfully leverage AI for academic growth.