Enhanced Imputation Methods Using Bivariate Auxiliary Information in Stratified Sampling: An application on Road Accident Data
Year: 2025
Type: article
Abstract: Non-response is a substantial obstacle to data quality and estimation precision in survey sampling, requiring appropriate ways to mitigate this problem. However, studies on imputation techniques utilizing bivariate auxiliary information within the framework of stratified simple random sampling (StSRS) are limited. We present novel combined imputation techniques that leverage bivariate auxiliary information through three strategies in StSRS to enhance population mean estimation. We formulate mean squared error (MSE) equations for both existing and novel methodologies and assess their efficacy via empirical and simulation analyses. Utilizing factory production and road accident data, we illustrate that the suggested estimators consistently attain reduced MSE and enhanced relative efficiency in comparison to existing methodologies. A simulation of 20,000 iterations further validates how well they perform across diverse non-response rates. These findings underscore the capability of dual auxiliary information to improve population parameter estimation, representing a notable development in survey sampling methodologies. Mathematical classification 62D05, 62D10.
Source: Alexandria Engineering Journal
Authors: Faizan Danish, G.R.V. Triveni, Umer Mushtaq Lone
Institution: VIT-AP University
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,900
OpenAlex ID: W4408178755