Estimation of Population Mean in Two-Phase Stratified Sampling

In this paper, we have considered the problem related to estimation of population mean under stratified random sampling in presence of non-response and measurement error. Two different chain exponential to regression type estimators are formed separately for different structures of two-phase stratified sampling schemes. To reduce nuisance effect of random non responses in practical surveys, we have developed effective imputation techniques. The properties of the estimates are studied. The performances of the proposed strategy have been demonstrated through numerical evidences carried over the data set of natural population and population generated through simulation studies. We have also examined the performances of proposed strategy for different values of random non-response probabilities. Suitable recommendations have also been put forward to the survey statisticians for their application in real life.