Biometrical Letters Vol. 49(1), 2012, pp. 1-36
|
ON THE GENERAL GAUSS-MARKOV MODEL FOR EXPERIMENTS IN BLOCK DESIGNS
Tadeusz Caliński Department of Mathematical and Statistical Methods, Poznań University of Life Sciences, Wojska Polskiego 28, 60-637 Poznań, e-mail: calinski@up.poznan.pl |
Main estimation and hypothesis testing results related to the Gauss-Markov model, in its general form, are recalled and the application of these results to the analysis of experiments in block designs is considered. Special attention is given to the randomization-derived model for a general block design, and for a proper block design in particular. The question whether the randomization-derived model can be considered as a particular general Gauss-Markov model is discussed. It is indicated that the former, as a mixed model, is in fact an extension of the general Gauss-Markov model. Thus, the analysis based on the randomizationderived model requires a more extended methodical approach. The present paper has been inspired by one of the last papers of Professor Wiktor Oktaba.
Block designs, estimation, general Gauss-Markov model, hypothesis testing, mixed model, randomization-derived model.
