<![CDATA[Central Java Province is one of the provinces with a higher percentage of poor population compared to the national average. Although the poverty rate has declined in recent years, poverty issues in Central Java still require comprehensive solutions, particularly through evidence-based, area-based policy formulation. This study aims to identify the existence of spatial dependence in the percentage of poor population in Central Java Province, to map local poverty clusters, and to detect statistically significant concentrations of high and low poverty values in order to reveal structural poverty pockets. Global spatial autocorrelation testing using Moran's I yielded a value of 0.181 (Z = 1.802; p = 0.036), indicating a statistically significant clustered spatial pattern in the percentage of poor population across regencies/cities. Local spatial autocorrelation analysis (LISA) further identified three types of local association patterns at the 10% significance level: High–High, Low–High, and High–Low. Complementing these local patterns, hotspot and coldspot analysis using the Getis–Ord statistic—which captures the concentration of high or low values within a neighborhood rather than deviation-based association—identified seven regencies as statistically significant hotspots (one at the 99%, three at the 95%, and three at the 90% confidence level) and one regency as a significant coldspot (95% confidence level). This dual approach enables a more specific and operational identification of priority regions for poverty reduction policy.]]>
Copyrights © 2026
