A new procedure for the analysis of early generation variety trials
A method for the analysis of early generation variety trials is described. The method is an extension of the analysis proposed by Gleeson and Cullis for replicated trials and uses the residual maximum likelihood estimation method of Patterson and Thompson. Best linear unbiased predictors of test line effects are derived. A small simulation experiment is conducted to assess the reliability of the method. A wheat trial with a weed density covariate and missing values is analysed to illustrate the method.