Bayesian and Likelihood Inference for the Generalized Fieller—Creasy Problem

  • M. Yin
  • M. Ghosh
Part of the Lecture Notes in Statistics book series (LNS, volume 148)


The famous Fieller-Creasy problem involves inference about the ratio of two normal means. It is quite challenging from either a frequentist or a likelihood perspective. Bayesian analysis with noninformative priors usually provides ideal solutions for this problem. In this paper, we find a second order matching prior and a one at a time reference prior which work well for Fieller-Creasy problem in the more general setting of two location-scale models with smooth symmetric density functions. The properties of the posterior distributions are investigated for some particular cases including the normal, t, and the double exponential. The Bayesian procedure is implemented via Markov Chain Monte Carlo (MC 2). Our simulation study indicates that the second order matching priors, in general, perform better than the reference priors in terms of matching the target coverage probabilities in a frequentist sense.


Posterior Distribution Coverage Probability Probability Match Profile Likelihood Likelihood Inference 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

7 References

  1. Barnard, G.A. and D.A. Sprott (1983). The generalised problem of the Nile: Robust confidence sets for parametric functions. Ann. Statist. 11, 104–113.MathSciNetMATHCrossRefGoogle Scholar
  2. Barndorff-Nielsen, O.E. (1983). On a formula for the distribution of the maximum likelihood estimator. Biometrika 70, 343–365.MathSciNetMATHCrossRefGoogle Scholar
  3. Berger, J.O. and J.M. Bernardo (1989). Estimating a product of means: Bayesian analysis with reference prior. J. Amer. Statist. Assoc. 84, 200–207.MathSciNetMATHCrossRefGoogle Scholar
  4. Berger, J.O. and J.M. Bernardo (1992). On the development of reference priors (with discussion). In J. Bernardo, J. Berger, A. Dawid, and A. Smith (Eds.), Bayesian Statistics IV, pp. 35–60. Oxford University Press.Google Scholar
  5. Berger, J.O., B. Liseo, and R. Wolpert (1999). Integrated likelihood methods for eliminating nuisance parameters. Statist. Sci. 14, 1–22.MathSciNetMATHCrossRefGoogle Scholar
  6. Bernardo, J.M. (1977). Inferences about the ratio of normal means: A. Bayesian approach to the Fieller-Creasy problem. In Recent Developments in Statistics, Proceedings of the 1976 European Meeting of Statisticians, pp. 345–350.Google Scholar
  7. Bernardo, J.M. (1979). Reference posterior distributions for Bayesian inference (with discussion). J. Roy. Statist. Soc. Ser. B 41, 113–147.MathSciNetMATHGoogle Scholar
  8. Carlin, B.P. and T.A. Louis (2000). Bayes and Empirical Bayes Methods for Data Analysis. London: Chapman and Hall.MATHCrossRefGoogle Scholar
  9. Chib, S. and E. Greenberg (1995). Understanding the Metropolis-Hastings algorithm. Amer. Statist. 49, 327–335.Google Scholar
  10. Cox, D.R. and N. Reid (1987). Parameter orthogonality and approximate conditional inference (with discussion). J. Roy. Statist. Soc. Ser. B 49, 1–39.MathSciNetMATHGoogle Scholar
  11. Creasy, M.A. (1954). Limits for the ratio of means (with discussion). J. Roy. Statist. Soc. Ser. B 16, 186–194.MathSciNetMATHGoogle Scholar
  12. Datta, G.S. and M. Ghosh (1995). Some remarks on noninformative priors. J. Amer. Statist. Assoc. 90, 1357–1363.MathSciNetMATHCrossRefGoogle Scholar
  13. Datta, G.S. and M. Ghosh (1996). On the invariance of noninformative priors. Ann. Statist. 24, 141–159.MathSciNetMATHCrossRefGoogle Scholar
  14. Fieller, E.C. (1954). Some problems in interval estimation (with discussion). J. Roy. Statist. Soc. Ser. B 16, 175–185.MathSciNetMATHGoogle Scholar
  15. Ghosh, M. and R. Mukerjee (1998). Recent developments on probability matching priors. In S. Ahmed, M. Ahsanullah, and B. Sinha (Eds.), Applied Statistical Science III. Nonparametric Statistics and Related Topics, pp. 227–252. Nova Science Publ., Inc.Google Scholar
  16. Gleser, L.J. and J.T. Hwang (1987). The nonexistence of 100(1 — α)% confidence sets of finite expected diameters in error-in-variable and related models. Ann. Statist. 15, 1351–1362.MathSciNetMATHCrossRefGoogle Scholar
  17. Johnson, R.A. (1970). Asymptotic expansions associated with posterior distributions. Ann. Math. Statist. 41, 851–864.MathSciNetMATHCrossRefGoogle Scholar
  18. Kappenman, R.F., S. Geisser, and C.F. Antle (1970). Bayesian and fiducial solutions to the Fieller-Creasy problem. J. Roy. Statist. Soc. Ser. B 32, 331–340.MathSciNetGoogle Scholar
  19. Liseo, B. (1993). Elimination of nuisance parameters with reference priors. Biometrika 80, 295–304.MathSciNetMATHCrossRefGoogle Scholar
  20. McCullagh, P. and R. Tibshirani (1990). A simple method for the adjustment of profile likelihood. J. Roy. Statist. Soc. Ser. B 52, 325–344.MathSciNetMATHGoogle Scholar
  21. Mendoza, M. (1996). A note on the confidence probabilities of reference priors for the calibration model. preprint.Google Scholar
  22. Mukerjee, R. and D.K. Dey (1993). Frequentist validity of posterior quan-tiles in the presence of a nuisance parameter: Higher order asymptotics. Biometrika 80, 499–505.MathSciNetMATHCrossRefGoogle Scholar
  23. Mukerjee, R. and M. Ghosh (1997). Second order probability matching priors. Biometrika 84, 970–975.MathSciNetMATHCrossRefGoogle Scholar
  24. Phillipe, A. and C.P. Robert (1994). A note on the confidence properties of reference priors for the calibration model, preprint.Google Scholar
  25. Reid, N. (1995). Likelihood and Bayesian approximation methods (with discussion). In J. Bernardo, J. Berger, A. Dawid, and A. Smith (Eds.), Bayesian Statistics, Volume 5, pp. 611–618. Oxford University Press.Google Scholar
  26. Sendra, M. (1982). Reference posterior distribution for the Fieller-Creasy problem. Test 33, 55–72.MathSciNetMATHGoogle Scholar
  27. Stephens, D.A. and A.F.M. Smith (1992). Sampling-resampling techniques for the computation of posterior densities in normal means problems. Test 1, 1–18.MathSciNetMATHCrossRefGoogle Scholar
  28. Tibshirani, R.J. (1989). Non-informative priors for one parameter of many. Biometrika 76, 604–608.MathSciNetMATHCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2001

Authors and Affiliations

  • M. Yin
  • M. Ghosh

There are no affiliations available

Personalised recommendations