Academic Self-Efficacy as a Mediator of the Effect of Personalized Learning on Academic Achievement among Chinese University Students

Abstract

This study examined academic self-efficacy as a mediator of the effect of personalized learning on academic achievement among graduate students in China. A quantitative cross-sectional survey design was employed. Data were collected from 450 graduate students in China through an electronic questionnaire, and 446 valid responses were retained after Mahalanobis distance screening. The research instrument consisted of 30 items measuring personalized learning, academic self-efficacy, and perceived academic achievement. Data were analysed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The measurement model showed satisfactory reliability and validity, with outer loadings above .70, Cronbach’s alpha values ranging from .930 to .934, composite reliability values ranging from .941 to .944, and AVE values ranging from .615 to .628. The structural model explained 66.0% of the variance in academic achievement and 57.4% of the variance in academic self-efficacy. The results showed that personalized learning had a significant positive direct effect on academic achievement (β = .329, t = 6.413, p < .001) and academic self-efficacy (β = .758, t = 31.583, p < .001). Academic self-efficacy also significantly predicted academic achievement (β = .534, t = 10.856, p < .001). Mediation analysis confirmed that academic self-efficacy significantly mediated the relationship between personalized learning and academic achievement (β = .405, t = 10.115, p < .001). The findings suggest that personalized learning improves graduate students’ academic achievement not only by providing flexible and adaptive learning support, but also by strengthening students’ confidence in their academic abilities.