JOURNAL ARTICLE

Determinan Angka Kematian Bayi di Jawa Barat Menggunakan Geographically Weighted Panel Regression

Abstract

Abstract. Panel data regression is a statistical method used to analyzing datasets that combine cross-section and time series dimensions. This approach facilitates the investigation of relationships between variables while accounting for temporal variations. However, in contexts where geographical factors play a significant role, conventional panel regression models often fail to adequately capture spatial heterogeneity across regions. To address this limitation, the Geographically Weighted Panel Regression (GWPR) model is employed. GWPR enables the estimation of location-spesific parameters through spatial weighting schemes. In this study, GWPR is applied to examine the determinants of infant mortality rates across districts and municipalities in West Jawa Province. The GWPR model utilizing a fixed exponential kernel yields the highest R-Squared value when compared to the standard panel regression model based on random effects. The results reveal that the factors influencing infant mortality rates vary across regions, with the proportion of low birth weight infants and the coverage of iron supplement distribution to pregnant women identified as the most consistent and statistically significant determinants. Abstrak. Regresi data panel merupakan metode statistik yang digunakan untuk menganalisis data yang menggabungkan unit cross-section dan time series. Metode ini memungkinkan analisis hubungan antar variabel dengan mempertimbangkan variasi yang terjadi dalam jangka waktu tertentu. Namun, dalam beberapa kasus, seperti data yang berkaitan dengan aspek geografis, regresi data panel standar tidak mampu menangkap perbedaan karakteristik antar wilayah secara optimal. Oleh karena itu, pendekatan Geographically Weighted Panel Regression (GWPR) digunakan untuk mengatasi heterogenitas spasial dalam data panel. GWPR memungkinkan estimasi parameter berbeda untuk setiap lokasi berdasarkan pembobotan spasial. dalam penelitian ini, GWPR diterapkan untuk mengetahui determinan angka kematian bayi (AKB) di Kabupaten/Kota Jawa Barat menggunakan GWPR. Model GWPR dengan pembobot fixed kernel exponential menghasilkan R-Square tertinggi dibandingkan dengan model regresi panel Random Effect Model. Hasil estimasi menunjukkan bahwa variabel yang berpengaruh terhadap AKB berbeda di setiap wilayah, dengan persentase bayi berat badan lahir rendah dan ibu hamil penerima tablet tambah darah menjadi faktor yang paling konsisten dan signifikan berpengaruh.

Keywords:
Geography

Metrics

0
Cited By
0.00
FWCI (Field Weighted Citation Impact)
0
Refs
0.27
Citation Normalized Percentile
Is in top 1%
Is in top 10%

Topics

Economic Growth and Fiscal Policies
Social Sciences →  Economics, Econometrics and Finance →  Economics and Econometrics
Livestock Farming and Management
Life Sciences →  Agricultural and Biological Sciences →  Agronomy and Crop Science
© 2026 ScienceGate Book Chapters — All rights reserved.