Constructing median-unbiased estimators in one-parameter families of distributions via stochastic ordering
Volume 30 / 2003
Applicationes Mathematicae 30 (2003), 251-255
MSC: 62F10, 62G05.
DOI: 10.4064/am30-3-1
Abstract
If $\theta \in { \Theta }$ is an unknown real parameter of a given distribution, we are interested in constructing an exactly median-unbiased estimator $\widehat \theta $ of $\theta $, i.e. an estimator $\widehat \theta $ such that a median $\textrm{Med}(\widehat \theta )$ of the estimator equals $\theta $, uniformly over $\theta \in { \Theta }$. We shall consider the problem in the case of a fixed sample size $n$ (nonasymptotic approach).
