On the Performance of Generalized Regression Estimator for Small Domains

Abstract

This article describes different types of indirect methods of estimation for small domains such as synthetic ratio estimation, synthetic regression estimation, composite estimation combining direct ratio with synthetic ratio estimation, generalized regression estimation and direct ratio estimation. We also compared the performances of the corresponding estimators in terms of absolute relative bias (ARB) and simulated relative standard errors (SRSE), which are computed by means of a simulation study of the crop acreage for the Inspector Land Revenue Circles (ILRCs) of state Rajasthan, India, considered as small domains. Further we consider the calibration approach to estimation for small domains and compare the performance of these estimators.