ENHANCING THE MEAN RATIO ESTIMATORS FOR ESTIMATING THE POPULATION

MEAN USING MIXED PARAMETERS

Damisa S.A., 2Alhaji B.B. and Doguwa S.I.S.

Department of Statistics, Ahmadu Bello University, Zaria

Department of Mathematics and Computer Science, Nigeria Defence Academy, Kaduna

Abstract

In this study we have incorporated a mixed parameter of the auxiliary variable to

propose three new improved ratio estimators to enhance the mean ratio estimators for

estimating the population mean. To enhance the efficiency of the proposed ratio

estimators, a linear combination of the coefficient of kurtosis, first decile, quartile

deviation, median and the population variance of the auxiliary variables have been

harnessed. The properties related to the suggested estimators are assessed through

constant, bias and mean square error. We have also provided empirical study for

illustration and corroboration using two population data in evaluating the

efficiencies of the proposed modified ratio estimators using percentage relative error

(PRE). The suggested ratio estimators performed better than the other existing ratio

estimators compared with in the study by emerging with a lesser mean squared error

and should be preferred in application over the existing estimators, due to their

smaller values of bias and Mean Square Error.

Keywords: Ratio, Estimation, Deciles, Variance, kurtosis, MSE