Biometrical Letters vol. 43(1), 2006, pp. 39-47
|
EFFECTIVENESS OF RESOLVABLE INCOMPLETE BLOCKS AND EXTENT OF SPATIAL RELATIONSHIPS IN VARIETY TESTING TRIALS ON WHEAT Kamila Tomaszyk, Wiesław Pilarczyk Department of Mathematical and Statistical Methods, Agricultural University of Poznań, Wojska Polskiego 28, 60-637 Poznań, e-mail: wpilar@au.poznan.pl |
In the paper the effectiveness of resolvable incomplete block designs in variety testing trials on wheat is reported. The results of 122 trials conducted for variety testing purposes are used. The relative (average) variance of treatment comparisons is used as a measure of the effectiveness of incomplete blocks. The same trial data are also used in investigations of spatial relationships between measurements. Ten different models for semivariance were fitted and compared. Parameters of the best fitted model (in every single trial) were used to describe the features of spatial relationship between trial data.
effectiveness, geo-statistics, incomplete blocks, semivariance
