PREDICTORS OF EDUCATIONAL RESILIENCE IN FLOOD-PRONE PUBLIC SCHOOLS OF NASEERABAD DIVISION: A HIERARCHICAL MODEL

Abstract

Flood-related disruption increasingly challenges schooling continuity in Pakistan, yet actionable school-level evidence on predictors of post-disruption recovery remains limited in many flood-prone districts. This study operationalized educational resilience as a continuity and recovery capability and tested a layered predictor model in public schools of Naseerabad Division, Balochistan. A cross-sectional questionnaire was administered to 80 schools (61.3% rural). Schools reported a mean of 7.29 closure days during the last flood event. Five multi-item indices were computed (structural resilience, institutional capacity, community engagement, psychosocial support capacity, and educational resilience), all showing high internal consistency (alpha range 0.955 to 0.971). Pearson correlations indicated that educational resilience aligned most strongly with structural resilience (r = 0.48) and institutional capacity (r = 0.49). In hierarchical regression, structural resilience explained 23.2% of variance in educational resilience, adding institutional capacity increased explained variance to 34.2% (delta R2 = 0.111, p = 0.001). Community engagement did not add a significant increment (delta R2 = 0.014, p = 0.200). Psychosocial support capacity contributed a further significant increment, yielding a final R2 of 0.397 (delta R2 = 0.040, p = 0.029). In the final model, structural resilience (beta = 0.37), institutional capacity (beta = 0.28), and psychosocial support capacity (beta = 0.21) were significant predictors. Findings support a sequenced planning approach that prioritizes safe-operational conditions and preparedness governance, while integrating psychosocial readiness to stabilize re-engagement after reopening.