An Investigation of The Efficiencies and Cost Effectiveness of Estimators of Population Mean in Two Occasion Successive Sampling Under Non-Response Scenario

Abstract

This study compares the performances of existing estimators of population mean in current occasion in two-occasion successive sampling under non-response scenario. Using dataset from Sharma et al. (2022), the estimators are evaluated across varying correlation structures, non-response rates, and optimum replacement policy. The results reveal distinct differences in efficiency among the estimators. Some estimators outperform other estimators, particularly when the correlations are strong. The percent relative loss (PRL) analysis further indicates that some estimators maintain stable performances even as non-response rates increase, demonstrating robustness under adverse survey conditions. Optimal unmatched proportions also contributed to significant gains in precision. Overall, the findings indicate that properly constructed estimators remain highly efficient under non-response and provide clearer guidance for selecting optimal estimators in repeated surveys.