Publication

Abstract : As technology continues to transform education, emerging realms of learning, engagement, and empowerment enabled by Augmented Reality (AR) and Large Language Models (LLMs) are only beginning to be uncovered. This study investigates their effects on student learning outcomes, self-efficacy, and cognitive load while mapping these impacts to Sustainable Development Goal 4 (SDG 4: Quality Education), contributing to a deeper understanding of their potential in shaping equitable and effective educational experiences. A total of 4,003 participants were involved, with 769 teachers (N= 246) and students (N = 523) in Part I assessing AR acceptance and 3,234 students in Part II comparing traditional passive classrooms, AR-based learning, and LLM-supported AR-based learning. Pre and post-test assessments measured conceptual understanding, self-efficacy, and cognitive load. The findings indicate that AR-based interventions significantly improve conceptual understanding compared to traditional classroom learning (p < 0.05). Mean values across SDG 4.3–4.7 indicators confirmed that AR and LLM integration in classrooms fosters higher engagement, comprehension, and accessibility. Mediation analysis further reveals that self-efficacy plays a crucial role in enhancing learning outcomes across all cognitive constructivist factors (p < 0.05), while cognitive load must be optimized to prevent learning inefficiencies. Notably, gender disparities observed in traditional classrooms were minimized in AR and LLM intervention, suggesting a more equitable learning environment. The results highlight AR’s strong contribution to SDG 4 by promoting equitable, high-quality, and inclusive education. While LLMs offer potential benefits, their role in AR learning requires further exploration. This study underscores the need for innovative, technology-driven pedagogical approaches to improve student learning and engagement, paving the way for future research on LLM-enhanced AR interventions in education.