ROBUST RATIO TYPE ESTIMATORS IN SIMPLE RANDOM SAMPLING USING HUBER M ESTIMATION
Keywords:
Quartiles, Non-Conventional location parameters, Median, M-Estimation, Bias, Mean Square Error, Efficiency.Abstract
In survey sampling a very dominating problem is to obtain the better ratio estimators of the population mean and population variance. A very striking question, in every researcher’s mind, is that if there are outliers in the data how to estimate the population mean and variance, as outliers mislead the results. So keeping the above problem under consideration, we propose some new modified ratio estimators using robust regression, which are robust against the outliers and give accurate results even in the presence of the outliers; also the properties are studied. It has been shown that the proposed class of estimators is more efficient than the existing classes of estimators. An empirical study has been carried out to examine the merits of the proposed class of estimators over others
