Improved Ratio-Regression Exponential Estimators For Population Mean Using Dual Auxiliary Variables Across Simple Random Sampling

Abstract

This article introduces improved estimators for assessing the population mean within the frame- work of simple random sampling (SRS), which incorporate two concomitant variables. Seven re- fined estimators are developed, and their mean [1]squared error (MSE) is derived using linear approximation. The proficiency of these estimators is evaluated through real-world datasets and simulation studies. Their performance is benchmarked against existing preliminary estimators using MSE and relative efficiency in the form of percentages as key metrics. Empirical analysis confirms the theoretical findings, showing that the new estimators outperform existing alternatives. Both the theoretical and empirical finding indicates that the proposed estimators achieve reduced mean squared error and improved percentage relative efficiency (PRE) when evaluated against conventional approaches.