{"id":"arxiv/2611.05600","url":"https://arxiv.org/abs/2611.05600","tags":["stat.CO","math.ST","stat.ME"],"score":0,"title":"Geometric ergodicity of the block Gibbs sampler for Bayesian linear mixed models with scale mixtures of normal random effects","domain":"stat.CO","source":"arxiv","authors":["Rolando de la Cruz","Nathalie Humeniy"],"reviews":0,"subject":"stat.CO","summary":"We study a Bayesian linear mixed model whose random effects follow a scale mixture of normal distributions with an arbitrary mixing distribution, a class that contains the Student-t, slash, contaminated normal, normal-generalized inverse Gaussian and Laplace laws. Conditionally conjugate priors lead to a three-block Gibbs sampler that alternates between the variance components, the latent precision weights, and the regression coefficients together with the random effects. Using a drift function built from fractional powers, we give sufficient conditions for geometric ergodicity of this…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-11-03","reactions":{}}