Dian Handayani, Khairil Anwar Notodiputro, Asep Saefuddin, I. Wayan Mangku, Anang Kurnia
Information about some poverty indicators is important not only for the large administrative level but also for lower administrative level. This information can be obtained from many surveys. Unfortunately, many surveys are usually designed to satisfy accuracy for large populations. As a result, it is often encountered that the sample size from some sub-populations which can be obtained from a survey is too small to produce a reliable direct estimator. The sub-population which the selected sample from it is not large enough to produce a reliable direct estimator is also called a small area. In this paper, we propose the spatial empirical best predictor (SEBP) for some poverty indicators in some small areas. The SEBP is derived under a unit-level spatial lognormal mixed model which incorporates spatial dependence into the covariance structure. The mean square prediction error (MSPE) of the SEBP is estimated by the parametric bootstrap method. A simulation study was conducted to evaluate the performance of the SEBP compared to the direct estimates as well as the empirical best predictor (EBP). Further, the SEBP was also applied to obtain the estimates of some poverty indicators for some sub-districts in Bogor, Indonesia. The results showed that there is a substantial reduction in MSPE of the SEBP over the direct estimates and the EBP for almost all sub-districts. © Al-Zaytoonah University of Jordan (ZUJ).
Department of Statistics, Universitas Negeri Jakarta, Indonesia; Department of Statistics, IPB University, Indonesia; Department of Mathematics, IPB University, Indonesia
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock