Estimating the mean value of occupational exposures.

It is very important to provide an accurate estimate for the mean value of lognormal distributed occupational exposures. Four commonly used methods for estimating a lognormal mean are the sample mean, the maximum likelihood estimate (MLE), a bias-corrected MLE, and the minimum variance unbiased estimator (MVUE). In this article the explicit expressions are given for the mean square errors of these four estimators, and performances of these four estimators are compared in terms of their mean square errors. This article reaffirms the conclusion of earlier researchers that the MVUE is uniformly superior to the other three estimators.