Resumen de Posterior Distributions for Func-tions of Variance Components

Irwin Guttman, Ulrich Menzefricke

• This paper uses a Bayesian approach to examine the posterior distributions of linear and non-linear functions of variance components in balanced two-way analysis of variance designs. The paper shows how exact posterior distributions for such functions can be readily described using Monte Carlo simulations where independent draws from the distributions are obtained. It is thus very straightforward to obtain percentiles, means, standard deviations, and other summary measures for the posterior distributions.