Is Bayes posterior just quick and dirty confidence?

DOI10.1214/11-STS352zbMATH Open1246.62040arXiv1112.5582OpenAlexW3101847394MaRDI QIDQ449832FDOQ449832

Authors: D. A. S. Fraser 

Publication date: 1 September 2012

Published in: Statistical Science (Search for Journal in Brave)

Abstract: Bayes [Philos. Trans. R. Soc. Lond. 53 (1763) 370--418; 54 296--325] introduced the observed likelihood function to statistical inference and provided a weight function to calibrate the parameter; he also introduced a confidence distribution on the parameter space but did not provide present justifications. Of course the names likelihood and confidence did not appear until much later: Fisher [Philos. Trans. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. 222 (1922) 309--368] for likelihood and Neyman [Philos. Trans. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. 237 (1937) 333--380] for confidence. Lindley [J. Roy. Statist. Soc. Ser. B 20 (1958) 102--107] showed that the Bayes and the confidence results were different when the model was not location. This paper examines the occurrence of true statements from the Bayes approach and from the confidence approach, and shows that the proportion of true statements in the Bayes case depends critically on the presence of linearity in the model; and with departure from this linearity the Bayes approach can be a poor approximation and be seriously misleading. Bayesian integration of weighted likelihood thus provides a first-order linear approximation to confidence, but without linearity can give substantially incorrect results.

Full work available at URL: https://arxiv.org/abs/1112.5582

Recommendations

• scientific article; zbMATH DE number 1067781
• On the Construction of Bayes–Confidence Regions
• Bayes and likelihood calculations from confidence intervals
• Confidence, Likelihood, Probability
• Frasian inference

zbMATH Keywords

posteriorpriorsBayes error ratedefault priorevaluating priorsnonlinear parameters

Mathematics Subject Classification ID

Bayesian inference (62F15) Foundations and philosophical topics in statistics (62A01) Parametric tolerance and confidence regions (62F25)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Outline of a Theory of Statistical Estimation Based on the Classical Theory of Probability
• Title not available (Why is that?)
• An invariant form for the prior probability in estimation problems
• Some Problems Connected with Statistical Inference
• Asymptotic Inference for Mixture Models by Using Data-Dependent Priors
• Second order ancillary: a differential view from continuity
• Title not available (Why is that?)
• Higher accuracy for Bayesian and frequentist inference: large sample theory for small sample likelihood
• An essay towards solving a problem in the doctrine of chances. By the late Rev. Mr. Bayes, F. R. S. communicated by Mr. Price, in a letter to John Canton, A. M. F. R. S.
• Title not available (Why is that?)
• Credible and confidence sets for restricted parameter spaces
• Setting confidence intervals for bounded parameters
• Title not available (Why is that?)
• Likelihood centered asymptotic model exponential and location model versions
• The problem of low counts in a signal plus noise model.
• Computation of distribution functions from likelihood information near observed data
• Holomorphic variations of minimal disks with boundary on a Lagrangian surface
• Some remarks on conditional and unconditional inference for location-scale models

Cited In (45)

• Confidence distribution for the ability parameter of the Rasch model
• Posterior model consistency in variable selection as the model dimension grows
• Discussion: Foundations of statistical inference, revisited
• Discussion of ``On the Birnbaum argument for the strong likelihood principle
• Direct and approximately valid probabilistic inference on a class of statistical functionals
• Improved inference on risk measures for univariate extremes
• On Some Principles of Statistical Inference
• Confidence, credibility and prediction
• On Bayesian credible sets, restricted parameter spaces and frequentist coverage
• A Bayesian tour of binomial inference
• When should modes of inference disagree? Some simple but challenging examples
• On the discrepancy between Bayes credibility and frequentist probability of coverage
• Accuracy, confidence and consensus in bayesian hypothesis inference
• Confidence distributions and empirical Bayes posterior distributions unified as distributions of evidential support
• The expected demise of the Bayes factor
• Controlling the degree of caution in statistical inference with the Bayesian and frequentist approaches as opposite extremes
• Bayes, reproducibility and the quest for truth
• Revisiting the Berger location model: fallacious confidence interval or a rigged example?
• D. A. S. Fraser: From structural inference to asymptotics
• Maximum entropy derived and generalized under idempotent probability to address Bayes-frequentist uncertainty and model revision uncertainty: an information-theoretic semantics for possibility theory
• Discussion of ``Is Bayes posterior just quick and dirty confidence? by D. A. S. Fraser
• Frasian inference
• “Not Only Defended But Also Applied”: The Perceived Absurdity of Bayesian Inference
• A fiducial approach to the nonparametric deconvolution problem: the discrete case
• The sufficiency of the evidence, the relevancy of the evidence, and quantifying both with a single number
• A comparison of learning rate selection methods in generalized Bayesian inference
• The Need for More Emphasis on Prediction: A “Nondenominational” Model-Based Approach
• Gibbs posterior inference on multivariate quantiles
• H‐likelihood Predictive Intervals for Unobservables
• Model-free posterior inference on the area under the receiver operating characteristic curve
• Confidence distributions: a review
• On default priors and approximate location models
• False confidence, non-additive beliefs, and valid statistical inference
• Confidence distributions and related themes
• Confidence is epistemic probability for empirical science
• On an inferential model construction using generalized associations
• Confidence distributions for change-points and regime shifts
• Limitations of proposed signatures of Bayesian confidence
• Interval estimation, point estimation, and null hypothesis significance testing calibrated by an estimated posterior probability of the null hypothesis
• The case of the Jeffreys-Lindley-paradox as a Bayes-frequentist compromise: a perspective based on the Rao-Lovric-theorem
• The \(p\)-value interpreted as the posterior probability of explaining the data: applications to multiple testing and to restricted parameter spaces
• Model misspecification, Bayesian versus credibility estimation, and Gibbs posteriors
• Inferential Models: A Framework for Prior-Free Posterior Probabilistic Inference
• Self-consistent confidence sets and tests of composite hypotheses applicable to restricted parameters
• The evidence interval and the Bayesian evidence value: On a unified theory for Bayesian hypothesis testing and interval estimation

This page was built for publication: Is Bayes posterior just quick and dirty confidence?

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q449832)