Vol: 35 Page: 7

Authors:
Henryk Brzeskwiniewicz
Jolanta Krzyszkowska

Title:
 Experiment design in group divisible block designs in relation to block size

Language: English

Keywords:
balanced block designs
group divisible block designs
basic contrasts
elementary contrasts
treatment contrasts
efficiency factor for treatment contrasts
Kronecker product of matrices
two-factorial experiments

Summary:
This paper presents two methods of experiment planning in group divisible block designs. The first method takes into consideration the empirical relation between the variance of experimental errors and the size of blocks, the second classical method does not take this relation into account. Examples show a great difference in the selection of an optimal design in both methods.