Measuring multidimensional student engagement beyond western contexts: A mixed-methods validation study in Mongolian higher education

Abstract

Student engagement is a multidimensional construct critical for academic success in higher education, yet most measurement instruments were developed in Western contexts and lack validation for non-Western settings, particularly regarding agentic engagement. This study adapted and validated a multidimensional student engagement instrument for Mongolian university students using a four-phase explanatory sequential mixed-methods design. The initial 38-item instrument combined Lam et al.'s (2014) Student Engagement Scale and Reeve and Tseng's (2011) Agentic Engagement Scale. Following translation, exploratory factor analysis (EFA; N = 216) identified 16 problematic items. Targeted cognitive interviews (N = 9) traced these item failures to contextual ambiguity, cultural inappropriateness, and construct reconceptualization. Evidence-based item revision was followed by confirmatory factor analysis (CFA; N = 275) with an independent sample. CFA supported a 20-item, four-factor model with excellent fit (χ²(163) = 247.89, p < .001; CFI = .968; TLI = .962; RMSEA = .043; SRMR = .057), strong reliability (subscale α = .839–.877; total α = .918), and adequate convergent and discriminant validity. Notably, agentic engagement emerged as a psychometrically sound and empirically distinct dimension despite Mongolia's traditionally teacher-centered pedagogical culture, challenging deficit-oriented assumptions about student agency in non-Western contexts. The relatively independent nature of behavioral engagement suggests that attendance-based participation may be normatively driven rather than reflective of deeper cognitive or affective involvement. The validated scale addresses a critical measurement gap in Mongolian higher education and offers a replicable mixed-methods validation framework applicable to other non-Western, resource-limited contexts.