Evaluation of Successive Sampling Strategy for Distribution Function Under Cost Function and Nonresponse Scenarios

Successive sampling is widely used in repeated surveys in which information on the same characteristic is collected overtime. In such designs, nonresponse and budget constraints substantially affect the performance of the estimator particularly for the distribution function (DF). This study proposed an improved class of unbiased estimator for the DF under two successive occasions. In practice, the values of the study variable evolve over time, and observations recorded on one occasion may not yield sufficient information for subsequent occasions. To address this challenge, we examine four distinct scenarios for estimating the finite population DF across two successive occasions in the presence of a nonresponse and cost function. The theoretical findings are validated through four real data sets and extensive simulation studies under normal and gamma distributions with varying sample size and nonresponse rates. The proposed framework is useful for statistical agencies, policy makers, and researchers. However, this study is limited to simple random sampling and first-order approximation under linear cost function.