Enhancing Population Variance Estimation through Log-Type Models in Neutrosophic Statistics

Accepted January 2026

Abstract

In traditional sampling theory, data are typically regarded as deterministic, lacking any uncertainty in the measurement of the studied attributes. Nonetheless, in real applications, data ambiguity is often encountered. In classical statistics, the population variance is computed using precise, definitive data values when auxiliary information is available. These estimates may frequently exhibit bias. This study’s primary goal is to address the drawbacks of traditional statistics while handling ambiguous data by introducing the neutrosophic estimator with the minimal mean squared error (MSE) for the unknown value of the population variance. In this study, the difficulty of measuring population variance is solved by introducing auxiliary data into the neutrosophic paradigm. Under the first-order approximation, the mean squared error (MSE) of the suggested log-type class of estimators is calculated. An empirical analysis based on two populations is carried out to evaluate their effectiveness, allowing for a useful comparison between the proposed log-type estimators and current estimators found in the literature.