Objective Priors: An Introduction for Frequentists

Abstract

Bayesian methods are increasingly applied in these days in the theory and practice of statistics. Any Bayesian inference depends on a likelihood and a prior. Ideally one would like to elicit a prior from related sources of information or past data. However, in its absence, Bayesian methods need to rely on some "objective" or "default" priors, and the resulting posterior inference can still be quite valuable. Not surprisingly, over the years, the catalog of objective priors also has become prohibitively large, and one has to set some specific criteria for the selection of such priors. Our aim is to review some of these criteria, compare their performance, and illustrate them with some simple examples. While for very large sample sizes, it does not possibly matter what objective prior one uses, the selection of such a prior does influence inference for small or moderate samples. For regular models where asymptotic normality holds, Jeffreys' general rule prior, the positive square root of the determinant of the Fisher information matrix, enjoys many optimality properties in the absence of nuisance parameters. In the presence of nuisance parameters, however, there are many other priors which emerge as optimal depending on the criterion selected. One new feature in this article is that a prior different from Jeffreys' is shown to be optimal under the chi-square divergence criterion even in the absence of nuisance parameters. The latter is also invariant under one-to-one reparameterization.

Published in at http://dx.doi.org/10.1214/10-STS338 the Statistical Science (http://www.imstat.org/sts/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (1)

• The formal definition of reference priors

Cited by in corpus (20)

• Inferential models: A framework for prior-free posterior probabilistic inference
• Overall Objective Priors
• Bayesian inference and the parametric bootstrap
• Multinomial Models with Linear Inequality Constraints: Overview and Improvements of Computational Methods for Bayesian Inference
• A note on p-values interpreted as plausibilities
• On an inferential model construction using generalized associations
• Uninformed Bayesian Quantum Thermometry
• Objective Bayesian inference with proper scoring rules
• Macroscopicity of quantum mechanical superposition tests via hypothesis falsification
• Quantum-classical hypothesis tests in macroscopic matter-wave interferometry
• Frameworks for prior-free posterior probabilistic inference
• Optimal post-selection inference for sparse signals: a nonparametric empirical-Bayes approach
• Congenial Differential Privacy under Mandated Disclosure
• A mathematical characterization of confidence as valid belief
• Optimality of Thompson Sampling with Noninformative Priors for Pareto Bandits
• Fundamental Issues Regarding Uncertainties in Artificial Neural Networks
• Matching Bayesian and frequentist coverage probabilities when using an approximate data covariance matrix
• Quantifying uncertainty in spikes estimated from calcium imaging data
• A default prior for regression coefficients
• There is Individualized Treatment. Why Not Individualized Inference?