Bayesian Analysis

Full Robustness in Bayesian Modelling of a Scale Parameter

Alain Desgagné

Full-text: Open access


Conflicting information, arising from prior misspecification or outlying observations, may contaminate the posterior inference in Bayesian modelling. The use of densities with sufficiently heavy tails usually leads to robust posterior inference, as the influence of the conflicting information decreases with the importance of the conflict. In this paper, we study full robustness in Bayesian modelling of a scale parameter. The log-slowly, log-regularly and log-exponentially varying functions as well as log-exponential credence (LE-credence) are introduced in order to characterize the tail behaviour of a density. The asymptotic behaviour of the marginal and the posterior is described and we find that the scale parameter given the complete information converges in distribution to the scale given the non-conflicting information, as the conflicting values (outliers and/or prior’s scale) tend to 0 or , at any given rate. We propose a new family of densities defined on R with a large spectrum of tail behaviours, called generalized exponential power of the second form (GEP2), and its exponential transformation defined on (0,), called log-GEP2, which proves to be helpful for robust modelling. Practical considerations are addressed through a case of combination of experts’ opinions, where non-robust and robust models are compared.

Article information

Bayesian Anal., Volume 8, Number 1 (2013), 187-220.

First available in Project Euclid: 4 March 2013

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Bayesian robustness conflicting information log-exponentially varying functions log-regularly varying functions log-slowly varying functions LE-credence Log-GEP_2


Desgagné, Alain. Full Robustness in Bayesian Modelling of a Scale Parameter. Bayesian Anal. 8 (2013), no. 1, 187--220. doi:10.1214/13-BA808.

Export citation


  • Andrade, J. A. A. and O’Hagan, A. (2006). “Bayesian robustness modeling using regularly varying distributions.” Bayesian Analysis, 1: 169–188.
  • Andrade, J. A. A. and O’Hagan, A. (2011). “Bayesian robustness modelling of location and scale parameters.” Scandinavian Journal of Statistics, 38: 691–711.
  • Angers, J.-F. (2000). “P-credence and outliers.” Metron, 58: 81–108.
  • Dawid, A. P. (1973). “Posterior expectations for large observations.” Biometrika, 60: 664–667.
  • De Finetti, B. (1961). “The Bayesian approach to the rejection of outliers.” In Proceedings of the 4th Berkeley Symposium on Mathematical Statistics and Probability, volume 1, 199–210. Berkeley: University of California Press.
  • Desgagné, A. and Angers, J.-F. (2005). “Importance sampling with the generalized exponential power density.” Statistics and Computing, 15: 189–195.
  • — (2007). “Conflicting information and location parameter inference.” Metron, 65: 67–97.
  • Karamata, J. (1930). “Sur un mode de croissance régulière des fonctions.” Mathematica (Cluj), 4: 38–53.
  • O’Hagan, A. (1979). “On outlier rejection phenomena in Bayes inference.” Journal of the Royal Statistical Society, Series B, 41: 358–367.
  • — (1988). “Modelling with heavy tails.” In Bernardo, J. M., DeGroot, M. H., Lindley, D. V., and Smith, A. F. M. (eds.), Bayesian Statistics 3, 345–359. Oxford: Clarendon Press.
  • — (1990). “Outliers and credence for location parameter inference.” Journal of the American Statistical Association, 85: 172–176.