From interest to engagement: a serial mediation model of self-efficacy and academic buoyancy

Abstract

Background International students learning Turkish as a Second Language face significant psychological challenges including linguistic anxiety, motivational fluctuations, and cultural adaptation stress that threaten sustained engagement. Despite the critical role of psychological resources in language acquisition, how motivational beliefs and emotional resilience interact to sustain engagement remains underexplored. This study examines the psychological mechanisms through which interest, academic self-efficacy, and academic buoyancy-the capacity to manage everyday academic setbacks-predict classroom engagement among culturally diverse learners. Methods Guided by Social Cognitive Theory, Expectancy-Value Theory, and positive psychology frameworks, the research tested a serial mediation model with 348 international students across 28 mother tongues in Turkish language centers. Participants ranged from 17 to 39 years (mean = 21.66 years). Psychological constructs were measured via validated scales: Academic Buoyancy Scale (resilience to routine challenges), Academic Efficacy Scale (competence beliefs), Language Classroom Engagement Scale (behavioral, cognitive, and emotional dimensions), and Interest Scale (intrinsic motivation). Structural equation modeling with maximum likelihood robust estimation examined psychological pathways. Bias-corrected bootstrapping with 5000 samples tested mediation effects. Results The structural model demonstrated excellent fit (Comparative Fit Index = 0.976, Tucker-Lewis Index = 0.972, Root Mean Square Error of Approximation = 0.037, Standardized Root Mean Square Residual = 0.032). Interest strongly predicted self-efficacy (standardized coefficient = .805, p < .001), revealing how intrinsic motivation builds competence beliefs. Self-efficacy significantly predicted buoyancy (standardized coefficient = .417, p < .001), demonstrating that confidence fosters psychological resilience. Buoyancy predicted sustained engagement (standardized coefficient = .283, p < .001). The serial mediation pathway was significant (standardized coefficient = .095, 95% confidence interval [0.033, 0.178]). These psychological mediators explained 47.5% of the total effect (standardized coefficient = .844), confirming that sustained engagement emerges through coordinated psychological processes. Conclusions This study advances motivational psychology in language acquisition by demonstrating a sequential pathway where interest catalyzes self-efficacy, which builds resilience, ultimately sustaining engagement. Psychological resources operate as an integrated system, with nearly half of motivational effects transmitted through this developmental chain. The findings extend positive psychology to culturally diverse contexts and indicate generalizability across 28 linguistic backgrounds. Psychological interventions must adopt holistic approaches targeting the entire motivational-resilience pathway to support learners facing cognitive and emotional challenges.