PERBANDINGAN MODEL REGRESI BINOMIAL NEGATIF DENGAN MODEL GEOGRAPHICALLY WEIGHTED POISSON REGRESSION (GWPR) (Studi Kasus : Angka Kematian Ibu di Provinsi Jawa Timur Tahun 2011)

Abstract

Maternal mortality rate is one of the crucial problems of death in Indonesia. Maternal deaths in East Java province is likely to increase so that the role of data and information are very important. Negative Binomial Regression is a model that can be used to address the problem overdispersion. While the method of spatial attention factor for type discrete data is Geographically Weighted Poisson Regression Model (GWPR). This study was conducted on the comparison between the Negative Binomial Regression and GWPR to discuss the factors that influence maternal mortality rate in the province of East Java. Indicators that affect maternal mortality include maternal health services. Maternal health services such as antenatal care, obstetric complications treated, Aid deliveries by skilled health care child birth, and neonatal health care services handled neonatal complications. The results of testing the suitability of model shows that there is no influence of spatial factors on maternal mortality rate in the province of East Java. Based on Negative Binomial Regression derived variable number of puerperal women who received vitamin A significantly affect maternal mortality rate, while for GWPR is divided into six clusters districts/cities by same significant variables. From the comparison value of AIC was found that GWPR better to analyzing Maternal mortality in East Java because it has the smallest value of AIC. Key word : Maternal mortality rate, Discrete, Overdispersion, Negative Binomial Regression, Geographically Weighted Poisson Regression, AIC