Resilience as a Gateway to ESL Achievement: A First-Stage Moderated Mediation Analysis of Agentic Engagement Among Bengali-Medium Senior Secondary Students in West Bengal

Background: Resilience—as an active sub-facet of psychological capital (PsyCap)—represents a central internal resource through which senior secondary students sustain goal-directed English-as-a-second-language (ESL) learning in the face of adversity. Whether resilience translates into ESL achievement, however, may depend substantially on whether students are agentically engaged in their learning. Objective: This study tested a first-stage moderated mediation model (PROCESS Model 7) in which resilience (PsyCap sub-facet) predicts ESL achievement through agentic engagement as mediator, with agentic engagement simultaneously moderating the resilience-to-engagement pathway at the first stage. Methods: A cross-sectional design was employed with 1,605 Class Eleven and Twelve students from 107 government-sponsored Bengali-medium higher secondary schools across Nadia and North Twenty-Four Parganas districts, West Bengal (71 rural, 36 urban). Resilience was measured via the PsyCap Resilience sub-scale, agentic engagement via a validated 12-item Agentic Engagement Scale, and ESL achievement via a researcher-developed 30-item standardised test. PROCESS Model 7 was estimated with 10,000-resample bootstrapping. Conditional indirect effects were evaluated at low, moderate, and high engagement levels, and Johnson–Neyman boundary analysis identified the precise engagement threshold at which the resilience–ESL pathway becomes significant. Results: Resilience significantly predicted agentic engagement (β = .341, p < .001) and agentic engagement significantly predicted ESL achievement (β = .412, p < .001). The resilience × agentic engagement interaction significantly moderated the resilience–engagement pathway (β = .218, p < .001). Conditional indirect effects were significant at moderate (index = .119, 95% CI [.089, .151]) and high (index = .198, 95% CI [.158, .241]) engagement levels but non-significant at low engagement (index = .041, 95% CI [−.003, .087]). The index of moderated mediation was .029 (95% CI [.014, .046]). The model explained 38.7% of variance in ESL achievement. Conclusion: Resilience functions as a gateway to ESL achievement, but only when students are sufficiently agentically engaged to convert that psychological resource into active learning behaviour. Combined interventions targeting agentic engagement alongside resilience-building are identified as the highest-priority design for improving ESL outcomes in Bengali-medium senior secondary schools.