Past and Present Statistical Issues in Design of Variety Trials

Abstract

In this article some past design issues are reviewed and some present issues of practical importance are highlighted. Lattice designs are common in plant breeding. Control varieties are included for comparison of varieties from different lattices. When the varieties are related by pedigree, the related varieties are sometimes grouped in consecutive plots. Alpha designs should be preferred in both these cases. In presence of interference, neighbour-balanced designs may improve the precision. Straw height differences could be used as a covariate. Spatial analysis may be more efficient than traditional block analysis. However, restricted randomisation is still needed. An exact theory for the relationship between design, randomisation and analysis is wanted

Published in

Title: Proceedings from the 13th International Conference and exhibition on mechnanization of field experiments

Conference

IAMFE 2008, The 13th International Conference and Exhibition on Mechanization of Field Experiments

SLU Authors

• Forkman, Johannes

  • Department of Crop Production Ecology, Swedish University of Agricultural Sciences
    

UKÄ Subject classification

Agricultural Science
Food Science


Permanent link to this page (URI)

https://res.slu.se/id/publ/18524