Prediction approach estimator using auxiliary information under two-phase sampling

Abstract

Sample survey is a cost effective mean to collect reliable information about a finite population. There are various sampling methodologies, among them two-phase sampling is generally used for estimating population mean or total under the two different situations. First, when the information of the auxiliary variable is not readily available and the other condition is when it is vey expensive to gather information on characteristic under study y, but it is comparatively cheaper to gather information on the variables which are highly correlated with the characteristic under study. In large scale surveys, two-phase sampling approach is proposed in order to reduce the number of sampled units which require the more expensive objective methods. Prediction approach is applied to predict the non-sampled units in surveys. In the large preliminary sample (first phase sample) of two-phase sampling, there are total n'- n non-sampled units having auxiliary information, so there is a need to develop an estimator based on prediction approach under finite population. In the present study, we have proposed a new estimator of finite population total based on prediction approach in the context of two-phase sampling.