Abstract
R. A. Fisher's conditionality principle is used to estimate the location parameter of an exponential distribution with known coefficient of variation. The minimal sufficient statistic includes an ancillary statistic and two estimators of the location parameter, based on the observed value of this ancillary statistic, are proposed. The conditional mean square errors of these estimators are compared with the corresponding unconditional mean square errors of the estimators obtained as optimum linear combinations of the components of the minimal sufficient statistic.
Get full access to this article
View all access options for this article.