Spatial inequalities in O/L to A/L school dropouts: A GIS and regression analysis of district-level patterns in Sri Lanka (2012–2024)

Abstract

Sri Lanka’s free education policy has produced high literacy and enrollment, yet dropout during the transition from the G.C.E. Ordinary Level (O/L) to Advanced Level (A/L) remains a persistent challenge. Each year, nearly 20,000 students exit the system without progressing to A/L, with dropouts concentrated in northern, plantation, and rural districts. While earlier studies link dropout to poverty and resource gaps, the spatial dimension of these disparities has rarely been examined systematically. This study aimed to investigate district-level disparities in O/L to A/L dropout between 2012 and 2024 and to assess the influence of educational and socio-economic resources. Guided by spatial inequality and educational stratification theories, dropout rates for 25 districts were mapped using ArcGIS Pro, with hotspot analysis identifying clusters. Explanatory variables, availability of 1AB schools, proportion of graduate-trained teachers, and relative wealth, were tested using multiple regression in Jamovi. Results showed dropout hotspots in Mullaitivu, Kilinochchi, Vavuniya, and Puttalam, while coldspots were observed in southern districts such as Matara and Galle. Nationally, relative wealth was the strongest negative predictor (p < 0.001, R² = 0.60), whereas teacher and school availability were weaker or inconsistent. In high-dropout districts, teacher availability correlated positively with dropout, suggesting reactive allocation, while no variable explained high-transition districts, indicating the role of cultural factors. The study concludes that dropout disparities are driven more by socio-economic inequality than infrastructure. Findings underscore the need for targeted interventions in disadvantaged districts and demonstrate the utility of combining GIS and regression for educational policy research.