MODELLING SCHOOL DROPOUT RATES IN WEST JAVA PROVINCE WITH MIXED GEOGRAPHICALLY TEMPORALLY WEIGHTED REGRESSION

Abstract

School dropout is a problem in the education sector that can hinder the progress of the quality of human resources and the competitiveness of the nation. West Java Province has the highest school dropout rates among all provinces in Indonesia. The data on school dropout rates exhibit spatial and temporal variations. Additionally, the potential differences between regions allow for the occurrence of diverse data that can be addressed locally and globally. Mixed Geographically Temporally Weighted Regression (MGTWR) is an extension of the GWR method that can produce parameters that are both local and global for each location and time. So, the objective of this research is to obtain factors that have a local and global influence on the school dropout rate in West Java Province using the Mixed Geographically Temporally Weighted Regression method. In this study, the data used includes school dropout rates in West Java Province from 2018 to 2022. The data used is sourced from the official statistical data website of the Ministry of Education, Culture, Research and Technology, and the official website of the West Java Province Central Statistics Agency. The results of the MGTWR modeling show that globally influential variables include the percentage of the poor population, population density, unemployment rate, and average length of schooling, which have local effects. Based on the MGTWR model, the Fixed Kernel Gaussian weighting function is the best model for modeling school dropout rates in regencies/cities in West Java, with an RMSE value of 0.0755 and R-squares of 92.09%.