This paper extends Besag's (1994) identifiability conditions to propose convergence conditions for the Gibbs sampler that are independent of the selected version of the conditional distributions. Moreover, we show that the support of the joint distribution must be connected if the Gibbs sampler...

We consider the construction of perfect samplers for posterior distributions associated with mixtures of exponential families and conjugate priors, starting with a perfect slice sampler in the spirit of Mira and co-workers. The methods rely on a marginalization akin to Rao-Blackwellization and...

In univariate calibration, two standard estimators are usually opposed: the classical estimator and the inverse regression estimator. Controversies have followed the use of both estimators and we consider them from a decision-theoretic perspective, establishing the inadmissibility of the...