Vol: 44 Page: 69

Authors:
Katarzyna Ambroży-Deręgowska
Iwona Mejza
Stanisław Mejza

Title:
 On the relative efficiency of split-split-plot design to split-plot × split-block design

Language: English

Keywords:
relative efficiency
split-split-plot design
split-plot x split-block design

Summary:
In the paper we consider two the most popular in practice designs for three-factorial experiments, i.e. split-split-plot (SSP) design and split–plot × split–block (SPSB) design. Discussed here models of observations are called randomized-derived models and are strictly connected with randomization performed for nested and crossed structures of experimental units. Statistical properties result from two different schemes of randomization applied in the experiments.
The aim of the paper is comparing both models with respect to the relative efficiency in order to decide which a design is the best for estimation and testing hypotheses for factors and interactions between them. The considerations are illustrated with an example involving a winter wheat trial.

Display