On measuring the variability of small area estimators under a basic area level model

摘要: SUMMARY In this paper based on a basic area level model we obtain second-order accurate approxi mations to the mean squared error of model-based small estimators, using Fay & Herriot (1979) iterative method estimating variance weighted residual sum squares. We also estimators unbiased second order. Based simulations, compare finite-sample performance our with those method-of-moments, maximum likelihood and variance. Our results suggest that Fay-Herriot performs better, in terms relative bias than other methods across different combinations number areas, pattern sampling variances distribution effects. derive noninformative prior parameters for which posterior is error. The such possesses both Bayesian frequentist interpretations.