<![CDATA[Stunting remains one of the most serious nutritional problems in Indonesia, with prevalence levels that vary considerably across provinces and over time. This study aims to identify the multidimensional determinants of stunting prevalence in Indonesia by accounting for both spatial and temporal heterogeneity. The analysis uses secondary panel data from thirty four provinces covering the period from two thousand twenty to two thousand twenty four, sourced from the Ministry of Health and the Central Statistics Agency. The method applied is Geographically Weighted Panel Regression, which integrates spatial and temporal dimensions to capture the diverse relationships between stunting and its determinants across regions. Panel model selection procedures indicated that the fixed effect specification best represented the data, while diagnostic tests confirmed the presence of significant spatial heterogeneity, justifying the use of a geographically weighted approach. Among the kernel weighting functions compared, the adaptive exponential kernel produced the best balance of model fit and accuracy. The findings reveal that the average length of schooling among women is the most dominant and consistent factor associated with lower stunting prevalence across many provinces. Female labor force participation shows a significant but spatially varying influence, while exclusive breastfeeding coverage is significant only in a limited number of regions. These results highlight that stunting reduction strategies should be tailored to the specific socioeconomic and spatial characteristics of each province rather than relying on a uniform national policy.]]>
