Centre for Statistical & Survey Methodology Working Paper Series

Publication Date



At a meeting held in Canberra in July 2010, it was agreed to undertake a study to investigate whether a combined analysis of the 08/09 series of LMA experiments and the single 2010 experiment was feasible. The current protocol for classification of lines relies on obtaining information from two valid LMA screening experiments. The analysis of the 08/09 series of experiments represented the first attempt to undertake a combined analysis of LMA screening experiments. It is well known (see for example, Mrode [1995]) that the optimum methodology to evaluate genetic material for eventual selection is to use a multivariate or multi-trait analysis. This multi-trait analysis takes account of the relationships which exist between the different traits. In our example, each experiment is considered as a trait and so for these data we have three traits, one for each experiment.