Perbandingan Geographically Weighted Regression dengan Mixed Geographically Weighted Regression
Studi Kasus Prevalensi Stunting di Indonesia
DOI:
https://doi.org/10.29313/statistika.v23i2.1700Keywords:
Geographically Weighted Regression , Mixed Geographically Weighted Regression, StuntingAbstract
ABSTRACT
Infants who suffer from chronic malnutrition, or stunting, have a very small height compared to other children their age. Slow-growing children have a higher risk of contracting diseases and growing up with degenerative conditions. Indonesia has the second highest stunting rate among children under five in ASEAN. In 2015, 36.4% of Indonesian children less than five years old were stunted. Indonesia has the second highest stunting rate in Southeast Asia, after Laos (43.8%). The child nutritional status study found that 29.6% of children under five years old were stunted (PSG, 2017). The number increased from 2017 to 2018. Stunting rates vary from province to province in Indonesia. Therefore, research linking stunting to the characteristics of the province is needed to address its prevalence. Therefore, there is a need for research that correlates stunting with relevant provincial characteristics to address its prevalence. By comparing the Geographically Weighted Regression (GWR) method and the Mixed Geographically Weighted Regression (MGWR) method, this study aims to identify the factors that influence the prevalence of stunting among under-fives in Indonesia by considering regional aspects in the province. The average incidence of stunting in Indonesia is 30.59. The study found that East Nusa Tenggara Province has the highest prevalence of stunting in Indonesia. In Indonesia, the success rate of children receiving all recommended vaccine doses (X1) was a significant predictor of stunting prevalence. The GWR model is superior to the MGWR model because the optimal AIC and R2 values are 167.6841 and 0.828, respectively, as shown by the comparison of global regression models.
ABSTRAK
Bayi yang menderita kekurangan gizi kronis, atau stunting, memiliki tinggi badan yang sangat kecil dibandingkan anak-anak lain seusianya. Anak-anak yang tumbuh lambat mempunyai risiko lebih tinggi tertular penyakit dan tumbuh dengan kondisi degeneratif. Indonesia merupakan negara dengan tingkat stunting tertinggi kedua di antara balita di ASEAN. Pada tahun 2015, 36,4% anak Indonesia berusia kurang dari lima tahun mengalami stunting. Indonesia merupakan negara dengan angka stunting tertinggi kedua di Asia Tenggara, setelah Laos (43,8%). Penelitian status gizi anak menemukan bahwa 29,6% balita mengalami stunting (PSG, 2017). Jumlah tersebut meningkat dari tahun 2017 ke tahun 2018. Tingkat stunting bervariasi dari satu provinsi ke provinsi lain di Indonesia. Oleh karena itu, diperlukan penelitian yang menghubungkan stunting dengan karakteristik provinsi terkait untuk mengatasi prevalensinya. Dengan membandingkan metode Geographically Weighted Regression (GWR) dan metode Mixed Geographically Weighted Regression (MGWR), penelitian ini bertujuan untuk mengidentifikasi faktor-faktor yang mempengaruhi prevalensi stunting pada balita di Indonesia dengan mempertimbangkan aspek regional di provinsi tersebut. Rata-rata kejadian stunting di Indonesia adalah 30,59. Penelitian ini menemukan bahwa provinsi Nusa Tenggara Timur memiliki prevalensi stunting tertinggi di Indonesia. Di Indonesia, tingkat keberhasilan anak mendapatkan seluruh dosis vaksin yang dianjurkan (X1) merupakan prediktor signifikan terhadap prevalensi stunting. Model GWR lebih unggul dibandingkan model MGWR karena nilai AIC dan R2 optimalnya masing-masing sebesar 167.6841 dan 0.828 yang ditunjukkan oleh hasil perbandingan model regresi global.
References
Anindita, Sabella Dinna, (2018). Pemodelan Presentase Balita Stunting di Indonesia Dengan Metode regresi Nonparameterik Spline Truncated”, Surabaya : Statistika, ITS.
Apriluana, G, 2018. Analisis Faktor- Faktor Risiko terhadap Kejadian Stunting pada Balita (0-59 Bulan) di negara Berkembang dan AsiaTenggara”.
Asnidar, A., Haerani, H., Sriyanah, N., & Suswani, A. (2022). Determinants of Stunting in Pre-School-Aged Children in Ujung Bulu Subdistrict. Proceedings of the International Conference on Nursing and Health Sciences, 3(1), 291-298. https://doi.org/10.37287/picnhs.v3i1.1233
Litbang Depkes:http://ejournal2.litbang.kemk es.go.id/index.php/mpk/article/downl oad/472/537.
Fadilah, Tawangki Sri & Kudus, Abdul, (2023). Penerapan Metode Regresi Kernel Smoothing untuk Imputasi Data Lama Waktu Terinfeksi Covid-19. https://doi.org/10.29313/jrs.v3i1.1802. Jurnal Riset Statistika (JRS). UNISBA.
Fotheringham, A.S., Brunsdon, C., & Charlton, M. (2002), Geographically Weighted Regression, Jhon Wiley & Sons, Chichester, UK
Islamiyah, I. (2022). Analysis Of Determinant Factors In Stunting Incidence In Toddlers. International Journal of Transdisciplinary Knowledge, 3(2), 35–42. https://doi.org/10.31332/ijtk.v3i2.31
Katadata, 2018. “Prevalensi Stunting Balita Indonesia Tertinggi Kedua di ASEAN”, Diakses dari halaman : Http:// databooks.katadata.co.id/datapublish/ 2018/11/22/prevalensi-stunting- balita-Indonesia-tertinggi-kedua-di- asean.
Katadata, 2019. “Cek Fakta, 1 dari 3 Balita di Indonesia Mengalami Stunting/ Kerdil” akses
https://databoks.katadata.co.id/data publish/2019/01/16/cek-fakta-1- dari-3-balita-di-indonesia- mengalami-stuntingkerdil
Kementrian Kesehatan Republik Indonesia, 2017. Buku Saku Pantauan Status Gizi (PSG) Tahun 2017. Diunduh dari http://www.kesmas.kemkes.go.id/.
Mei C. L., He S. Y., Fang K. T., (2004), “A note on the mixed geographically weighted regression model" Journal of Regional Science, 44, 143-157
Pusat Data dan Informasi Kementerian Kesehatan RI. (2018). Situasi Balita Pendek (Stunting) di Indonesia. Jakarta: Kementerian Kesehatan RI. World Bank. (2007). Comprehensive implementation plan on maternal, infant and young child nutrition. Geneva: WHO.
Yosza, Frandy. (2018). Kejadian Stunting di Indonesia. Diakses dari http : himaep.feb.unair.ac.id/thinking-out- cloud/128-stunting-di-indonesia.html