Online learning self-efficacy and self-regulated learning in relation to perceived cognitive and academic performance in digital learning: an explanatory sequential mixed-methods study at five Vietnamese universities

Abstract

This explanatory sequential mixed-methods study examined associations of self-regulated learning (SRL) and online learning self-efficacy with perceived academic and cognitive performance among undergraduates recruited from five Vietnamese universities. The quantitative phase analyzed 685 complete survey responses obtained through quota-guided convenience sampling and tested a four-construct covariance-based structural equation model. The qualitative phase involved 24 interview participants purposively selected from the survey pool to contextualize the observed structural associations. The measurement model showed good fit on the retained indices (χ²/df = 2.15, CFI = .962, TLI = .958, RMSEA = .041, SRMR = .038). SRL was positively associated with perceived academic performance ( β  = .51) and perceived cognitive performance (β = .44), while online learning self-efficacy was associated with perceived academic performance ( β  = .24) and perceived cognitive performance ( β  = .36), all p  < .001. The model accounted for 49.5% and 52.1% of the variance in the two outcomes, respectively. Interview accounts described SRL through scheduling, micro-deadlines, monitoring, distraction control, and strategy revision, whereas online learning self-efficacy was reflected in willingness to initiate tasks, explore unfamiliar tools, seek help, and persist through technological difficulty. The study contributes a theoretically integrated account of learner agency grounded in SRL theory, Social Cognitive Theory, and Self-Determination Theory. Because the design was cross-sectional and the sample was nonprobability-based, the findings are interpreted as sample-specific associations rather than causal or nationally representative effects.