An Improved Alternative Method of Imputation for Missing Data in Survey Sampling

Authors

G. N. Singh, Indian Institute of Technology (Indian School of Mines), Dhanbad, India - 826004Follow
Ashok. K. Jaiswal, Indian Institute of Technology (Indian School of Mines), Dhanbad, India - 826004Follow
Chandraketu Singh, Department of Statistics, CHRIST (Deemed to be University), Bengaluru, Karnataka, IndiaFollow
M. Usman, Indian Institute of Technology (Indian School of Mines), Dhanbad, India - 826004Follow

Abstract

In the present paper, a new and improved method of ratio type imputation and corresponding point estimator to estimate the finite population mean is proposed in case of missing data problem. It has been shown that this estimator utilizes the readily available auxiliary information efficiently and gives better results than the ratio and mean methods of imputation; furthermore, its efficiency is also compared with the regression method of imputation and some other imputation methods, discussed in this article, using four real data sets. A simulation study is carried out to verify theoretical outcomes, and suitable recommendations are made.

Digital Object Identifier (DOI)

http://dx.doi.org/10.18576/jsap/110214

Recommended Citation

N. Singh, G.; K. Jaiswal, Ashok.; Singh, Chandraketu; and Usman, M. (2022) "An Improved Alternative Method of Imputation for Missing Data in Survey Sampling," Journal of Statistics Applications & Probability: Vol. 11: Iss. 2, Article 14.
DOI: http://dx.doi.org/10.18576/jsap/110214
Available at: https://digitalcommons.aaru.edu.jo/jsap/vol11/iss2/14