Akademik Veri Yönetim Sistemi
Quality and Quantity, 2025 (Scopus)
This study introduces a novel generalized chain regression-cum-chain ratio estimator for estimating population means using two auxiliary variables within a robust two-phase sampling framework. The proposed estimator incorporates an innovative sub-sampling mechanism specifically designed to mitigate the effects of non-response—a prevalent issue in survey sampling. Three distinct non-response scenarios are examined: (i) non-response in the study variable with complete information on both auxiliary variables, (ii) non-response affecting one or both auxiliary variables, and (iii) non-response in a single auxiliary variable while the other remains fully observed. For each case, explicit expressions for bias and mean squared error (MSE) are derived to enable analytical comparisons with traditional unbiased and existing chain-type estimators. Theoretical findings are further validated through extensive simulation experiments and empirical applications. Results demonstrate that the proposed estimator consistently yields higher efficiency and lower estimation error, establishing it as a reliable and practical advancement for improving population mean estimation under non-response conditions.