Analysis of Variance Calculations for Irregular Experiments
Irregular experiments may be more useful than much smaller regular experiments and can be analyzed statistically without undue expenditure of time. For a few missing plots, standard methods of calculating missing-plot values can be used. For more missing plots (up to 10 percent), seedlot means or randomly chosen plot means of the same seedlot can be substituted for missing plots, provided between-block differences are small. Whatever the number of missing plots (provided there is more than one plot per seedlot) or the size of the between-block differences, seedlot means and sums of squares can be estimated in terms of deviations from block means. The procedures for calculation of analysis of variance in terms of deviations are described. The procedures are also applicable to regular experiments.
Download this file:
Download this file — PDF document, 140KbDetails
Author(s): Jonathan W. Wright
Publication: Tree Improvement and Genetics - Lake States Forest Tree Improvement Conference - 1977