An effective and economic estimation of population mean in stratified random sampling using a linear cost function

Abstract

Improvement in the estimation of population mean has been an area of interest in sampling theory. So many estimators have been suggested for elevated estimation of the population mean in stratified random sampling, but there is still a gap for more closely estimating the population mean. In this paper, the authors propose a ratio-product-cum-exponential-cum-logarithmic type estimator for the enhanced estimation of population mean by implying one auxiliary variable in stratified random sampling using conventional ratio, exponential ratio, and logarithmic ratio type estimators. The suggested estimator is a generalization of ratio, exponential ratio, and logarithmic ratio type estimators, and therefore these are special cases of the proposed estimator. The proposed estimator's bias and MSE are determined and compared with those of influential estimators, with the linear cost function being used to investigate and compare alternatives. Use Cramer's rule to determine the optimal value of the proposed estimator. The proposed estimator is more effective than other existing estimators, according to theoretical observations. For various applications, we suggest using a proposed estimator with the minimal MSE, which is verified by a numerical example, to have practical applicability of theoretical conclusions in real life.

Keywords

• Auxiliary variable
• Cramer's rule
• MSE
• PRE and linear cost function