Estimating the Mean of Finite Population under Double Sampling Stratification in the Presence of Non-Response

Abstract

Estimating finite population mean is the primary concern in many studies, particulary when the non response occurs for some units. This study focuses in estimating the finite population mean using the auxiliary variables under double sampling stratification (DSS) in the presence of non-response which occurs on both the study and the auxiliary variables. A Regression-ratio-in-ratio type exponential strategy is proposed with non-response simultaneously. Expressions for bias and mean square errors (MSE) are derived up-to-first order of approximation. MSE and percentage relative efficiency (PRE) are computed numerically using vrious real data sets. A simulation study is also conducted to verify the performance of estimators. Results indicate that the proposed estimator has the minimum MSE and the maximum PRE compared to competitor estimators. Therefore the proposed estimator is more efficient and is recommended for practical use in future.