Identifikasi Faktor-Faktor Penyebab Kejadian Diare Di Kota Semarang Dengan Pendekatan Geographically Weighted Poisson Regression

Abstract

The percentage of people affected by diarrheal diseases are still quite high, reaching 5.2%. Therefore we need an effort to identify the factors that cause diarrhea efforts of the government in order to reduce morbidity of diarrhea optimally. Such efforts include reviewing of the factors causing the incidence of diarrhea by focusing on linkages between regions or spatial aspects. Spatial aspect is considered important to study because between regions must have different characteristics. One approach that can be used is a spatial model Geographically Weighted Poisson Regression (GWPR) which is a local form of the Poisson Regression. This research was conducted in Semarang city with the unit of observation is the 16 districts in Semarang city. The results showed that the locally influential variable is the number of protected drinking water facilities and the number of medical personnel available. This model has a level of accuracy of 84.33%.