BIAS-CALIBRATED ESTIMATION FROM SAMPLE-SURVEYS CONTAINING OUTLIERS

We discuss the problem of estimating finite population parameters on t he basis of a sample containing representative outliers. We clarity th e motivation for chambers's bias-calibrated estimator of the populatio n total and show that bias calibration is a key idea in constructing e stimators of finite population parameters. We then link the problem of estimating the population total to distribution function or quantile estimation and explore a methodology based on the use of Chambers's es timator. We also propose methodology based on the use of robust estima tes and a bias-calibrated form of the Chambers and Dunstan estimator o f the population distribution function. This proposal leads to a bias- calibrated estimator of the population total which is an alternative t o that of Chambers. We present a small simulation study to illustrate the utility of these estimators.