Ojenike Oluwadamilola, T. And Ishaq O. Olawoyin

When the variance among different strata significantly surpasses the variance within each stratum, stratification becomes instrumental in boosting efficiency. A novel exponential estimator was introduced for stratified random sampling. Its validity was confirmed through application to datasets concerning the volume of Apple production during three distinct instances in 1999 within Turkey. Strata were defined by Turkish regions, and samples were randomly chosen from each region using the Neyman allocation method. Mathematical expressions for the mean squared error (MSE) of the proposed estimators were formulated up to the first-order approximation. Computational assessments of bias and MSE equations for the suggested estimator families demonstrated unbiasedness and efficiency in their MSE. Theoretical contrasts and numerical evaluations revealed the superiority of the proposed estimator over existing ones. Furthermore, the newly derived exponential estimator enhanced performance in estimating population means through stratified random sampling, especially when auxiliary attributes were considered. Keywords: Auxiliary attribute; Mean square errors (MSE); Percentage Relative Efficiency (PRE); Variance; Bias 0150