A study on new methods of ratio exponential type imputation in sample surveys

Authors : Shakti PRASAD

Pages : 1281-1301

View : 26 | Download : 8

Publication Date : 2018-10-16

Article Type : Research Paper

Abstract :In this article, we have suggested new methods of ratio exponential type  imputation and proposed their corresponding point estimators to deal with the problems of non-response in sample surveys for the prior outlay  of an auxiliary variable $x$. The expression of the biases and their mean square errors of the proposed estimators have been derived, upto the fi rst order of large sample approximation under SRSWOR scheme and compared with the mean method of imputation, ratio method of impu tation, regression method of imputation and the estimators of Singh and Horn insert ignore into journalissuearticles values(Metrika [16]);, Singh and Deo insert ignore into journalissuearticles values(Statistical Papers [15]);, Touten burg et al. insert ignore into journalissuearticles values(Statistical Papers [18]);, Singh insert ignore into journalissuearticles values(Statistics [17]); and Gira insert ignore into journalissuearticles values(Applied Mathematical Sciences [5]);. After comparison, the condition which makes the proposed forty four estimators more efficient than others are found. To verify the theoretical results, simulation studies are performed on five real data sets. 
Keywords : Eciency, Imputation methods, Bias, Mean square error MSE, Auxiliary variable