Advances in Regression Models used for Business Statistics

Abstract

The performance of an estimator is generally judged on the basis of relative variance, relative standard error or standard error. These measures are generally unknown since they are the function of the population parameter, thus estimated on the basis of sample information. The expressions of relative bias and relative variance of these estimators are obtained when sampling is done from a finite population and the sample size is large. It is also examined that which of these estimators be adopted as a reasonable criterion for judging the performance of an estimator. The results are verified in case of simple random sampling without replacement.