<![CDATA[Digital transformation in education demands continuous improvement of teacher competencies in coding and artificial intelligence (AI). The mechanism explaining why some trained teachers adopt digital innovations while others do not remains underexplored, particularly self-efficacy as a mediating pathway between training and adoption. This study analyzes post-training self-efficacy, its predictive effect on adoption intention, and whether digital competency mediates this relationship. A cross-sectional survey involved 62 teachers (TK/PAUD, elementary, junior high school) participating in a Coding and AI Technical Guidance (Bimtek) program organized by a Regional Education Office in Central Java, Indonesia, selected through total sampling. Three validated instruments were administered: a self-efficacy questionnaire (10 items, α = .904), an adoption intention questionnaire (8 items, α = .925), and a 15-item digital competency test. Data were analyzed using simple linear regression and the Baron and Kenny mediation procedure verified by the Sobel test, with post-hoc power computed using Cohen's f² benchmark. Technology self-efficacy was high (M = 4.11, SD = 0.56) and significantly predicted adoption intention (β = .776, R² = .602, p < .001), with achieved power exceeding .99 (f² = 1.51). Digital competency did not mediate this relationship (Sobel z = 0.408, p = .684), indicating self-efficacy operates through a direct affective-motivational pathway rather than competency accumulation. This study provides novel evidence that self-efficacy is a more proximal determinant of adoption intention than competency in short-duration training. Implications include redesigning Bimtek to target self-efficacy through mastery experiences and assessing self-efficacy as a training outcome indicator.]]>
Copyrights © 2026
