Admissible linear estimation in a general Gauss-Markov model with an incorrectly specified dispersion matrix (Q1116216)

Necessary and sufficient conditions are established for the set of all admissible linear estimators under \({\mathbb{M}}_ 0\) to be contained in the corresponding set of estimators under \({\mathbb{M}}\), where \({\mathbb{M}}_ 0\) and \({\mathbb{M}}\) are general Gauss-Markov models with identical model matrices but different dispersion matrices. As preliminary results, certain new characterizations of admissible linear estimators are derived, including explicit expressions for the general representations of such estimators and extensions of the admissibility criteria given by \textit{C. R. Rao} [Ann. Statist. 4, 1023-1037 (1976; Zbl 0336.62055)] and \textit{W. Klonecki} and \textit{S. Zontek} [J. Multivariate Anal. 24, 11-30 (1988)].