Zurück zu den Preprints des Jahres 2006


2006-42

On Optimal Designs in Random Intercept Models

by Thomas Schmelter, Rainer Schwabe.


Series: 2006-42, Preprints

MSC:
62K05 Optimal designs
62J05 Linear regression

Abstract:
In the one-sample case standard optimal designs retain their optimality even if a random intercept is present. However, in a multi-sample situation the variability of the intercept may substantially influence the choice of the optimal design.

Keywords:
optimal design, random intercept, random block effect, treatment comparison