REGRESI SPATIAL DURBIN MODEL UNTUK MENGIDENTIFIKASI FAKTOR YANG BERPENGARUH PADA ANGKA KEMATIAN BAYI DI JAWA TIMUR

Infant Mortality Rate (IMR) is an important indicator of health sector development that can describe the state of health status in a society. Among the provinces in Indonesia in 2009, the position of IMR in East Java, including the middle class that is equal to 31 deaths per 1000 live births. However there are quite high IMR variability between districts in East Java. This study aims to descriptions of the IMR and the factors that influence it in East Java from the territorial point of view, as well as modeling the IMR in East Java by using the Spatial Durbin Model. Aspects in this research area related to the characteristics and cultural differences between regions. The benefits are achieved is to give researcher information and references to particular government agencies to improve the quality of infant health as well as inputs to the government in taking a policy to reduce IMR in East Java. Source of research data was obtained from the publication of the macro socio-economic BPS in 2009. Analysis of the IMR is used for modeling regression methods Spatial Durbin model (SDM). The results showed that the distribution pattern of the IMR are grouping patterns. IMR was highest in the region and the island of Madura Pandalungan region (north coast of East Java Province). Parameter estimation results can be shown that the lagged dependent variable and independent influence on IMR modeling and factors that influence it. Variables that have a significant effect is the percentage of deliveries performed by medical assistance and the percentage of households using clean water. Key words : IMR, Spatial Durbin Model Abstrak. Angka Kematian Bayi (AKB) merupakan indikator penting dalam pembangunan sektor kesehatan yang dapat menggambarkan keadaan derajat kesehatan di suatu masyarakat. Di antara provinsi-provinsi di Indonesia pada tahun 2009, posisi AKB di Jawa Timur termasuk kelompok menengah yaitu sebesar 31 kematian per 1000 kelahiran hidup. Walau demikian terjadi variabilitas AKB yang cukup tinggi diantara kabupaten di Jawa Timur. Penelitian ini bertujuan bagaimana deskripsi AKB dan faktor-faktor yang mempengaruhinya di Jawa Timur dari sudut pandang kewilayahannya, serta memodelkan AKB di Jawa Timur dengan menggunakan metode Spatial Durbin Model. Aspek wilayah dalam penelitian ini terkait dengan perbedaan karakteristik dan kebudayaan antar daerah. Manfaat yang dicapai adalah memberi informasi dan referensi kepada peneliti dan instansi pemerintah guna meningkatkan kualitas kesehatan bayi serta sebagai masukan dalam mengambil kebijakan untuk menurunkan