SPATIAL DISTRIBUTION ANALYSIS OF ELEMENTARY SCHOOL POINTS USING GIS-BASED POISSON MODEL

Abstract

The equitable distribution of basic education facilities is an important indicator of fair regional planning. This study analyzes the spatial distribution pattern of Elementary School (SD) points in Bener Meriah Regency, Aceh Province, and models the number of schools against regional sub-district characteristics using a Geographic Information System (GIS) approach integrated with Poisson regression. School data were obtained from the Indonesian Ministry of Education's reference database (Dapodik), covering a full census of 143 SD-equivalent schools across 10 sub-districts, with a stratified random sub-sample of 34 geocoded schools used for point pattern analysis. Spatial pattern was assessed using Average Nearest Neighbor (ANN) analysis, while the effect of population on school count per sub-district was modeled using Poisson regression with area as an exposure offset. The ANN result yielded a ratio of R = 0.80 (z = -2.24; p < 0.05), indicating a statistically significant clustered pattern. The Poisson model across 10 sub-districts showed population to be a significant positive predictor of school count (IRR = 1.086 per 1,000 people; p < 0.001), although overdispersion was detected (deviance/df = 6.93), suggesting a Negative Binomial model for future work with a larger number of analysis units. These findings indicate spatial inequality in elementary school provision across sub-districts that warrants attention in education facility planning in Bener Meriah Regency.