Bayesian Analysis

The Evidentiary Credible Region

David Shalloway

Full-text: Open access

Abstract

Many disparate definitions of Bayesian credible intervals and regions are in use, which can lead to ambiguous presentation of results. It is particularly unsatisfactory when intervals are specified that do not match the one-sided character of the evidence. We suggest that a sensible resolution is to use the parameterization-independent region that maximizes the information gain between the initial prior and posterior distributions, as assessed by their Kullback-Leibler divergence, subject to the constraint on included posterior probability. This turns out to be equivalent to the relative surprise region previously defined by Evans (1997), and thus provides information theoretic support for its use. We also show that this region is the constrained optimizer over the posterior measure of any strictly monotonic function of the likelihood, which explains its many optimal properties, and that it is guaranteed to be consistent with the sidedness of the evidence. Because all of its equivalent derivations depend on the evidence as well as on the posterior distribution, we suggest that it be called the evidentiary credible region.

Article information

Source
Bayesian Anal. Volume 9, Number 4 (2014), 909-922.

Dates
First available in Project Euclid: 21 November 2014

Permanent link to this document
https://projecteuclid.org/euclid.ba/1416579184

Digital Object Identifier
doi:10.1214/14-BA883

Mathematical Reviews number (MathSciNet)
MR3293961

Zentralblatt MATH identifier
1327.62163

Keywords
credible region credible interval highest posterior density parameterization invariance Kullback-Leibler information gain relative surprise region

Citation

Shalloway, David. The Evidentiary Credible Region. Bayesian Anal. 9 (2014), no. 4, 909--922. doi:10.1214/14-BA883. https://projecteuclid.org/euclid.ba/1416579184.


Export citation

References

  • Baskurt, Z. and Evans, M. (2013). “Hypothesis assessment and inequalities for Bayes factors and relative belief ratios.” Bayesian Analysis, 8: 569–590.
  • Berger, J. O., Bernardo, J. M., and Sun, D. (2009). “The formal definition of reference priors.” Annals of Statistics, 37: 905–938.
  • Bernardo, J. (1979). “Reference posterior distributions for Bayesian inference.” Journal of the Royal Statistical Society. Series B (Methodological), 41: 113–147.
  • — (2005). “Intrinsic credible regions: An objective Bayesian approach to interval estimation.” Test, 14: 317–384.
  • Box, G. E. and Tiao, G. C. (1992). Bayesian inference in statistical analysis. Hoboken, NJ: Wiley-Interscience.
  • Casella, G. and Berger, R. L. (1990). Statistical Inference. Pacific Grove, CA: Wadsworth & Brooks.
  • Chen, M.-H. and Shao, Q.-M. (1999). “Monte Carlo estimation of Bayesian crediable and HPD intervals.” Journal of Computational and Graphical Statistics, 8: 69–92.
  • Evans, M. (1997). “Bayesian inference procedures derived via the concept of relative surprise.” Communications in Statistics - Theory and Methods, 26: 1125–1143.
  • Evans, M., Guttman, I., and Swartz, T. (2006). “Optimality and computations for relative surprise inferences.” Canadian Journal of Statistics, 34: 113–29.
  • Evans, M. and Shakhatreh, M. (2008). “Optimal properties of some Bayesian inferences.” Electronic Journal of Statistics, 2: 1268–280.
  • Good, I. (1988). “Surprise index.” In Kotz, S., Johnson, N., and Reed, C. (eds.), Encyclopedia of Statistical Sciences, volume 7. New York: John Wiley & Sons.
  • — (1989). “Surprise indices and p-values.” Journal of Statistical and Computational Simulation, 32: 90–92.
  • Nocedal, J. and Wright, S. (2006). Numerical Optimization. Springer Series in Operations Research and Financial Engineering. New York: Springer, second edition.
  • Turkkan, N. and Pham-Gia, T. (1997). “Highest posterior density credible region and minmum area confidence region: the bivariate case.” Applied Statistics, 46: 131–182.
  • Wei, G. and Tanner, M. (1990). “Calculating the content and boundary of the highest posterior density region via data augmentation.” Biometrika, 77: 649–652.