Abstract
Sample survey data can be used to derive reliable direct estimates for large areas or domains, but sample sizes in small areas or domains are seldom large enough for direct estimators to provide adequate precision for small areas. This makes it necessary to employ indirect estimators that borrow strength from related areas. We provide empirical Bayes (EB) or empirical best linear unbiased prediction (EBLUP) estimators under two basic models: unit level and area level. Methods for measuring the variability of the EB estimator are compared. Simple modifications to the estimator of mean squared error, proposed by (1990), are given. These estimators are area-specific, unlike the Prasad-Rao estimator, in the sense of depending on area-specific data.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
