Discussion of Analysis of variancewhy it is more important than ever by A. Gelman.pdfVIP

The Annals of Statistics 2005, Vol. 33, No. 1, 40–43 DOI: 10.1214/009053604000001048 c Institute of Mathematical Statistics, 2005 5 DISCUSSION OF “ANALYSIS OF VARIANCE—WHY IT IS MORE 0 0 IMPORTANT THAN EVER” BY A. GELMAN 2 By Joop Hox and Herbert Hoijtink g u A Utrecht University 6 2 Bayesian inference and fixed and random effects. Professor Gelman writes “Bayesians see analysis of variance as an inflexible classical method.” He ] adopts a hierarchical Bayesian framework to “identify ANOVA with the T S structuring of parameters into batches.” In this framework he sidesteps “the . overloaded terms fixed and random” and defines effects “as constant if they h t are identical for all groups in a population and varying if they are allowed a to differ from group to group.” Applying this approach to his first example m (a Latin square with five treatments randomized to a 5 × 5 array of plots), [ variance components have to be estimated for row, column and treatment 1 effects. v In our opinion, his approach provides an insightful connection between 8 2 analysis of variance and hierarchical modeling. It renders an informative and 5 easy to interpret display of variance components that is a nice alternative for 8 traditional analysis of variance. However, we wonder whether sidestepping 0 5