The estimated proportion of chronic disease sufferer in Duren Sawit district, East Jakarta, using hierarchical bayes method in small area estimation (SAE)

The purpose of this research is to observe the proportion of chronic disease sufferer in the Duren Sawit district, East Jakarta. The data that used in this research is the primary data in the form of survey data directly and the secondary data in the form of census data from Dinas Kesehatan (Dinkes) 2017 and Badan Pusat Statistik (BPS) 2017. The sampling method used is simple random sampling with a sample size of 1% of the total heads of families living in the Duren Sawit district, that is 1229 the heads of the family. On direct estimation, estimating parameter only based on survey data of subpopulations is inappropriate action, because the sample size obtained relatively few or there are subpopulations that is not selected as the sample. To overcome this, indirect estimation is used with small area estimation (SAE) method, which borrowed extra information such as administrative data or census data from other areas or area itself and there's an addition random area effect into the model. In this research, the proportion of chronic disease sufferer in Duren Sawit district is obtained used direct estimation and indirect estimation with hierarchical bayes method in SAE. The results of the estimates proportion from direct estimation is 18.87% with Mean Square Error (MSE) is 0.000130741, whereas the estimates of the proportion from indirect estimation is 18.37% with MSE 0.0000922915. Based on the MSE that obtained, the estimate proportion of chronic disease sufferer from indirect estimation is more reliable than direct estimation.