On the inadmissibility of unbiased estimators
It is observed that unbiased estimators are always inadmissible when the parameter (or function of the parameter) being estimated has either a maximum or a minimum at a parameter value for which the probability distribution is nondegenerate. Examples of problems where this is so include variance components problems, problems with restricted parameter spaces, and estimation of the risk or variance of shrinkage estimators.
Year of publication: |
1990
|
---|---|
Authors: | Berger, James O. |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 9.1990, 5, p. 381-384
|
Publisher: |
Elsevier |
Saved in:
Online Resource
Saved in favorites
Similar items by person
-
STATISTICAL DECISION THEORY: FOUNDATIONS, CONCEPTS, AND METHODS
Berger, James O., (1984)
-
BEHAVIOUR OF THE POSTERIOR DISTRIBUTION AND INFERENCES FOR A NORMAL MEAN WITH t PRIOR DISTRIBUTIONS
Fan, Tsai-Hung, (1992)
-
Fong, Duncan Κ. H., (1993)
- More ...