Cluster Information of Non-Sampled Area In Small Area Estimation

Abstract

Empirical Best Linear Unbiased Predictor (EBLUP) has been widely used to predict parameters in area with small or even zero sample size. The problem is when this model should be used to predict the parameters of non-sampled area. Ordinary EBLUP predicted the parameters using synthetic model which ignore the area random effects because lack of non-sampled area information. Thus, those prediction will be distorted based on a single line of the synthetic model. One of idea that developed in this paper is to modify the prediction model by adding cluster information by assuming that there are similiarities among particular areas. These information will be added into the model to modify the intercept of prediction model. Another approach is by adding random effects of auxiliary variable into the previous model in order to modify both intercept and slope of the prediction model. In this paper, simulation process is carried out to study the performance of the proposed models compared with ordinary EBLUP. All models evaluated based on the value of Relative Bias (RB) and Relative Root Mean Squares Error (RRMSE). The results show that the addition of cluster information can improve the ability of the model to predict on non-sampled areas.

Cluster information of non-sampled area in small area estimation of poverty indicators using Empirical Bayes

“…Cluster analysis on small area estimation can be used when there is a non-sampled area. The addition of cluster information to non-sampled areas shows that in general has a better prediction [3].…”

“…Yet, problems occur when this method is used to estimate parameter of nonsampled area which is solely based on synthetic model which ignore the area effects. A research by adding cluster information at standard model of EBLUP to substitute random effect of non-sampled area and the result showed that the addition of cluster information at standard model EBLUP gave better effect and estimates precision in non-sampled area [3]. Other research about best clustering methods showed that factor analysis with Ward methods is the best cluster [6].…”

“…Empirical Bayes methods in non-sampled area only based on synthetic model which ignore the area effects [1]. Anisa, Anang and Indahwati have researched the effect of adding the cluster information Empirical Unbiased Best Linear Prediction on non-sampled area and produce better predictions [3].…”

Clustering Information of Non-Sampled Area in Small Area Estimation of Poverty Indicators

“…Cluster analysis can be used when there is nonsampled area. The addition of cluster information to non-sampled areas shows that in general has a better prediction [4].…”

“…The problem of the ordinary EB for non-sampled area using synthetic model is ignoring the area effects. Anisa et al (2014) studied the effect of adding the cluster information empirical unbiased best linear prediction on non-sampled area to produce better predictions [4]. The simulation study showed the use of factor analysis in clustering has increased the average percentage of accuracy particularly when the Ward method is implemented [5].…”