Spatial Probit Regression Analysis Using Recursive Importance Sampling on the Human Development Index in Central Jawa Province

This research aims to analyze the Human Development Index (HDI) in Central Java Province using two modeling approaches: the probit model and the spatial probit model. HDI, a crucial indicator of quality of life, is categorized into two groups: medium-low and high. Among the 35 districts and cities in Central Java, 9 fall into the medium-low category (35%), while 26 are classified as high (65%). Brebes Regency has the lowest HDI at 69.54, whereas Sukoharjo City boasts the highest HDI at 77.73. The results reveal that the spatial probit model provides a superior classification of HDI compared to the probit model. The mapping of prediction results highlights the discrepancies between actual and predicted HDI categories. This study underscores the advantage of incorporating a spatial approach, enhancing the accuracy of HDI analysis, and offering a more nuanced understanding of spatial distribution and dependencies in human development within Central Java. Further research is recommended to include additional variables, such as poverty levels or access to education, which spatial factors may also affect. Moreover, this method can be applied to other regions or enhanced using more complex spatial models, such as spatial autoregressive (SAR) or geographically weighted regression (GWR), for deeper and more comprehensive analysis.