The integration of Artificial Intelligence (AI) into education continues to expand, yet its effectiveness in improving learning outcomes requires further examination, particularly with regard to the role of self-efficacy in digital learning in Indonesia. This study extends Bandura’s self-efficacy construct by conceptualizing AI as a latent mediating variable functioning as an adaptive educational mechanism through feedback, personalized resources, and dynamic scaffolding. A quantitative approach was employed via an online survey conducted between February and May 2024, yielding 279 valid responses from secondary and higher education students across several Indonesian provinces. Data were analyzed using SEM-PLS with SmartPLS 3.0. The results show that self-efficacy moderately predicts AI adoption (β = 0.373), AI integration strongly predicts learning behavior and performance (β = 0.649), while the direct pathway from self-efficacy is relatively weak (β = 0.144), with the indirect pathway mediated by AI substantially stronger (mediation effect = 0.242). Theoretically, these findings enrich Bandura’s framework of self-efficacy in the context of digital learning by highlighting AI as a central mediating construct, while practically they provide implications for adaptive learning strategies, digital education policy, and technology-driven pedagogical innovation.
Siregar et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: