Spatial modelling for understanding the correlation between school facilities and academic performance in the Philippines

Abstract

School facilities serve as places of learning for young people and are proven to have significant effect on education outcomes especially for developing countries. Since substantial funding is required for the provision of suitable school facilities, local context and limited resources in developing countries necessitate the prioritization of school facilities. The main purpose of this study is to examine the spatial variation of school facilities and identify which feature has the largest impact on academic performance in selected localities. The Philippines is chosen as a case study due to the recent reform in its education system and consideration of its unique geographic characteristics. Acquired government-provided and open-source school facility data as of 2013 have revealed that only 38% of its public school buildings are in good condition. This study employed statistical spatial analysis methods to understand the relationship between education, society and the environment. Regionalization and cluster analysis has revealed that school buildings in the capital are overcrowded but are in relatively better conditions than in the natural calamity-prone eastern seaboard region and in some southern provinces experiencing civil conflict. Global, local, and semi- parametric regression analyses indicate that the observed spatial variations in the provision of resources among the government schools are affecting academic performance. In the nation’s capital, public elementary schools primarily serve the urban poor. An examination of education facilities in its largest city indicates that the provision of school health clinics is correlated with better education outcomes. All in all, schools in urbanized areas in the Philippines should focus on the provision of large schools with lower pupil-teacher ratios. The provision of better school facilities had the most positive effect on academic performance in sparsely populated rural areas where the communities still lack these basic services. As the identified facilities influencing academic achievement varied depending on the location’s social and economic infrastructure, the provision of school facilities should be based on the unique needs of each community. Semi-parametric, geographically weighted regression modelling outperformed global and local modelling, and estimated up to 30% of the variation in education outcomes.