Abstract
For a scalar random-effect variance, Browne and Draper (2005) have found that the uniform prior works well. It would be valuable to know more about the vector case, in which a second-stage prior on the random effects variance matrix ${\bf D}$ is needed. We suggest consideration of an inverse Wishart prior for ${\bf D}$ where the scale matrix is determined from the first-stage variance.
Citation
Robert E. Kass. Ranjini Natarajan. "A default conjugate prior for variance components in generalized linear mixed models (comment on article by Browne and Draper)." Bayesian Anal. 1 (3) 535 - 542, September 2006. https://doi.org/10.1214/06-BA117B
Information