An improved class of synthetic estimator for mean of domain in missing data

Small area (domain) estimation in the case of missing data is a notable issue in sample surveys. Generally sample surveys are effected due to the phenomenon of missing data taking place due to various reasons, but estimates of small area are specifically more inclined to it. This paper is the result of the basic efforts made towards this issue. In the current paper, an improved class of synthetic estimator for mean of domain through imputation technique using simple random sampling without replacement (SRSWOR) has been proposed and its properties are examined. Some of the estimators available in literature are found as the special cases of the proposed estimator. Mathematical expressions for comparison are gotten to show the dominance of proposed estimator over the corresponding estimators. Then, it is extended through empirical analysis considering two authentic data sets. Further, simulation analysis is done to prove the results drawn through theoretical and empirical study.