Robust Spatial Autoregressive (Robust SAR) Modeling in the Case of Poverty Percentage in West Java

  • Novi Y
  • Tessy Octavia Mukhti
  • Zamahsary Martha
N/ACitations
Citations of this article
5Readers
Mendeley users who have this article in their library.

Abstract

Poverty is a complex problem influenced by various economic and social factors, such as the open unemployment rate, the minimum wage, population density, and the school participation rate. This study aims to model the poverty rate in West Java Province by considering spatial effects and the existence of outliers through the application of Spatial Autoregressive (SAR) and Robust Spatial Autoregressive (Robust SAR) models. Based on the Lagrange Multiplier test, the SAR model is declared suitable for use. However, the presence of outliers in the data necessitated the use of a robust approach to obtain more accurate results. The analysis showed that the Robust SAR model had a coefficient of determination  of 81.53%, higher than that of the SAR model at 77.48%, making it a better model for explaining variations in poverty levels. Of the four independent variables, only School Participation Rate had a significant effect in both models, where an increase in School Participation Rate contributed to a decrease in the poverty rate. This finding confirms the importance of investment in education as a strategic effort to reduce welfare inequality between regions in West Java.Kemiskinan merupakan masalah kompleks yang dipengaruhi oleh berbagai faktor ekonomi dan sosial, seperti tingkat pengangguran terbuka, upah minimum, kepadatan penduduk, dan tingkat partisipasi sekolah. Studi ini bertujuan untuk memodelkan tingkat kemiskinan di Provinsi Jawa Barat dengan mempertimbangkan efek spasial dan keberadaan outlier melalui penerapan model Spatial Autoregressive (SAR) dan Robust Spatial Autoregressive (Robust SAR). Berdasarkan uji Lagrange Multiplier, model SAR dinyatakan cocok untuk digunakan. Namun, keberadaan outlier dalam data mengharuskan penggunaan pendekatan robust untuk mendapatkan hasil yang lebih akurat. Analisis menunjukkan bahwa model Robust SAR memiliki koefisien determinasi sebesar 81,53%, lebih tinggi daripada model SAR sebesar 77,48%, sehingga menjadi model yang lebih baik untuk menjelaskan variasi tingkat kemiskinan. Dari empat variabel independen, hanya Tingkat Partisipasi Sekolah yang memiliki efek signifikan dalam kedua model, di mana peningkatan Tingkat Partisipasi Sekolah berkontribusi pada penurunan tingkat kemiskinan. Temuan ini menegaskan pentingnya investasi dalam pendidikan sebagai upaya strategis untuk mengurangi ketidaksetaraan kesejahteraan antar wilayah di Jawa Barat.

Cite

CITATION STYLE

APA

Novi, Y. M., Tessy Octavia Mukhti, & Zamahsary Martha. (2025). Robust Spatial Autoregressive (Robust SAR) Modeling in the Case of Poverty Percentage in West Java. Jurnal MSA ( Matematika Dan Statistika Serta Aplikasinya), 13(2), 123–133. https://doi.org/10.24252/msa.v13i2.61818

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free