Model Survival Spatial With Conditionally Autoregressive Frailty Pada Kasus Kematian Bayi Di Pulau Jawa

Prasetyo, Bayu (2017) Model Survival Spatial With Conditionally Autoregressive Frailty Pada Kasus Kematian Bayi Di Pulau Jawa. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
1315201709-Master-Theses.pdf - Published Version

Download (4MB) | Preview

Abstract

Angka kematian bayi di Indonesia yang masih tinggi akan menjadi tantangan dalam menghadapi Tujuan Pembangunan Berkelanjutan (SDGs). Untuk menurunkan angka kematian bayi diperlukan suatu pemahaman yang komprehensif tentang determinan kematian bayi termasuk laju kematian. Faktor-faktor yang diduga berpengaruh terhadap laju kematian pada bayi meliputi jenis kelamin, urutan kelahiran, penolong kelahiran, usia ibu saat kawin pertama dan saat melahirkan, pendidikan ibu serta akses air minum yang layak. Faktor perbedaan wilayah juga diduga memberi variansi dalam laju kematian. Penelitian ini menggunakan Bayesian Survival Spatial untuk menganalisis faktor-faktor yang mempengaruhi laju kematian pada bayi mati dibawah 1 tahun. Model menyertakan efek random/frailty spasial berdistribusi CAR (Conditionally Autoregressive), untuk menangkap variansi yang dihasilkan oleh korelasi spasial. Penelitian ini menggunakan matriks pembobot Queen’s contiguity dan Customized contiguity. Untuk mengetahui pengaruh ketetanggan antar wilayah terhadap kematian bayi, maka digunakan Statistik Uji Moran’s I yang menunjukkan nilai statistik Moran’s I sebesar 0.1394 dan nilai Z-value sebesar 2.2007 sehingga disimpulkan bahwa terdapat pengaruh spasial yang signifikan pada kematian bayi di setiap kabupaten/kota di Pulau Jawa. Distribusi weibull 2-parameter merupakan distribusi yang paling sesuai untuk memodelkan laju kematian. Faktor yang berpengaruh signifikan terhadap laju kematian bayi yaitu jenis kelamin bayi, urutan kelahiran bayi, penolong kelahiran bayi, usia ibu saat kawin pertama, usia ibu saat melahirkan bayi, ijazah tertinggi ibu, dan sumber air minum layak. Efek random mempengaruhi laju kematian bayi hanya pada komponen varian. ===================================================================================== Infant mortality rate in Indonesia that still high will be a challenge on Sustainable Development Goals (SDGs). An effort to reduce infant mortality rates requires a comprehensive understanding of the determinants of infant mortality, including hazard rates. The suspected factors influence the infant mortality rate are gender, birth order, birth attendants, maternal age at first marriage and childbirth, maternal education and access to decent drinking water. Differences of residential areas also allegedly gave the variation in hazard rate to death. This study uses a Bayesian Spatial survival to analyze the factors that affect hazard rate for infants under 1 year. This model includes the effects of random / CAR frailty (Autoregressive conditionally), to capture the variance was generated by spatial autocorrelation. This study uses a weighting matrix Queen's contiguity and Customized contiguity. To determine the effect of neighborhood between the regions to infant mortaity, is used the Test Statistics Moran's I, which shows the statistical value of Moran's I of 0.1394 and a Z-value of 2.2007 so its is concluded that there are significant spatial significant mortality baby in each regency/ city in Java. Weibull 2-parameter distribution is the most appropriate distribution to model the mortality rate. Significant variables that influence the rate of infant mortality are the baby's gender, birth order, birth attendants, maternal age at first marriage, maternal age when childbirth, maternal highest education, and decent drinking water sources. Random effects influence the hazard rate only at variance components.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Survival spatial; Frailty; MCMC; kematian bayi; Infant Mortality
Subjects: H Social Sciences > HA Statistics
H Social Sciences > HN Social history and conditions. Social problems. Social reform
Divisions: Faculty of Mathematics and Science > Statistics > (S2) Master Theses
Depositing User: - BAYU PRASETYO
Date Deposited: 11 Apr 2017 07:45
Last Modified: 08 Mar 2019 07:20
URI: http://repository.its.ac.id/id/eprint/3872

Actions (login required)

View Item View Item