Bemerkungen zum Vergleich linearer normaler Experimente. (Remarks on the comparison of linear normal experiments) (Q1106576)

The information comparison \({\mathcal E}_ A> {\mathcal E}_ B\) between two linear normal experiments \({\mathcal E}_ A\), \({\mathcal E}_ B\) is studied in terms of the Loewner ordering \(A\geq B\) between their defining design matrices A,B respectively. The results due to \textit{J. Kiefer} [J. Roy. Stat. Soc., Ser. B 21, 272-319 (ng-generated likelihoods. Saddlepoint approximations are shown to be easy to use and accurate in a variety of simple bootstrap and randomization applications. Examples include mean estimation, ratio estimation, two-sample comparisons, and autoregressive estimation.