A generalized marginal random effects model is described that enables exact Bayesian inference using either the Jeffreys or Berger-Bernardo non-informative prior distributions without the need for Markov Chain Monte Carlo ...