Bayesian Analysis

On the Existence of Uniformly Most Powerful Bayesian Tests With Application to Non-Central Chi-Squared Tests

Amir Nikooienejad and Valen E. Johnson

Advance publication

This article is in its final form and can be cited using the date of online publication and the DOI.

Full-text: Open access

Abstract

Uniformly most powerful Bayesian tests (UMPBT’s) are an objective class of Bayesian hypothesis tests that can be considered the Bayesian counterpart of classical uniformly most powerful tests. Because the rejection regions of UMPBT’s can be matched to the rejection regions of classical uniformly most powerful tests (UMPTs), UMPBT’s provide a mechanism for calibrating Bayesian evidence thresholds, Bayes factors, classical significance levels and p-values. The purpose of this article is to expand the application of UMPBT’s outside the class of exponential family models. Specifically, we introduce sufficient conditions for the existence of UMPBT’s and propose a unified approach for their derivation. An important application of our methodology is the extension of UMPBT’s to testing whether the non-centrality parameter of a chi-squared distribution is zero. The resulting tests have broad applicability, providing default alternative hypotheses to compute Bayes factors in, for example, Pearson’s chi-squared test for goodness-of-fit, tests of independence in contingency tables, and likelihood ratio, score and Wald tests.

Article information

Source
Bayesian Anal., Advance publication (2020), 17 pages.

Dates
First available in Project Euclid: 7 January 2020

Permanent link to this document
https://projecteuclid.org/euclid.ba/1578387712

Digital Object Identifier
doi:10.1214/19-BA1194

Keywords
uniformly most powerful Bayesian tests Bayesian hypothesis test chi-squared tests test of independence in contingency tables

Rights
Creative Commons Attribution 4.0 International License.

Citation

Nikooienejad, Amir; Johnson, Valen E. On the Existence of Uniformly Most Powerful Bayesian Tests With Application to Non-Central Chi-Squared Tests. Bayesian Anal., advance publication, 7 January 2020. doi:10.1214/19-BA1194. https://projecteuclid.org/euclid.ba/1578387712


Export citation

References

  • Albert, J. H. (1990). “A Bayesian test for a two-way contingency table using independence priors.” Canadian Journal of Statistics, 18(4): 347–363.
  • Bayarri, M. and García-Donato, G. (2008). “Generalization of Jeffreys divergence-based priors for Bayesian hypothesis testing.” Journal of the Royal Statistical Society: Series B (Statistical Methodology), 70: 981–1003.
  • Bayarri, M. J., Berger, J. O., Forte, A., García-Donato, G., et al. (2012). “Criteria for Bayesian model choice with application to variable selection.” The Annals of statistics, 40(3): 1550–1577.
  • Berger, J. O. and Pericchi, L. R. (1996). “The intrinsic Bayes factor for model selection and prediction.” Journal of the American Statistical Association, 91(433): 109–122.
  • Diaconis, P. and Lehmann, E. (2008). “Comment-On Student’s 1908 Article “The Probable Error of a Mean”.” Journal of the American Statistical Association, 103(481): 16.
  • Goddard, S. D. and Johnson, V. E. (2016). “Restricted most powerful Bayesian tests for linear models.” Scandinavian Journal of Statistics.
  • Good, I. and Crook, J. (1987). “The robustness and sensitivity of the mixed-Dirichlet Bayesian test for “independence” in contingency tables.” The Annals of Statistics, 670–693.
  • Hu, J. and Johnson, V. E. (2009). “Bayesian model selection using test statistics.” Journal of the Royal Statistical Society: Series B (Statistical Methodology), 71(1): 143–158.
  • Jeffreys, H. (1961). The theory of probability. Oxford Univ. Press, Oxford, 3rd edition.
  • Johnson, V. E. (2005). “Bayes factors based on test statistics.” Journal of the Royal Statistical Society: Series B (Statistical Methodology), 67(5): 689–701.
  • Johnson, V. E. (2013a). “Revised standards for statistical evidence.” Proceedings of the National Academy of Sciences, 110(48): 19313–19317.
  • Johnson, V. E. (2013b). “Uniformly most powerful Bayesian tests.” Annals of statistics, 41(4): 1716.
  • Johnson, V. E. and Rossell, D. (2010). “On the use of non-local prior densities in Bayesian hypothesis tests.” Journal of the Royal Statistical Society: Series B (Statistical Methodology), 72(2): 143–170.
  • Lehmann, E. L., Stein, C., et al. (1948). “Most powerful tests of composite hypotheses. I. Normal distributions.” The Annals of Mathematical Statistics, 19(4): 495–516.
  • Moreno, E., Bertolino, F., and Racugno, W. (1998). “An intrinsic limiting procedure for model selection and hypothesis testing.” Journal of the American Statistical Association, 93: 1451–1460.
  • Nikooienejad, A. and Johnson, V. E. (2020). “On the Existence of Uniformly Most Powerful Bayesian Tests With Application to Non-Central Chi-Squared Tests. Supplementary Material.” Bayesian Analysis.
  • O’Hagan, A. (1995). “Fractional Bayes factors for model comparison.” Journal of the Royal Statistical Society. Series B (Methodological), 99–138.
  • Patnaik, P. (1949). “The non-central $\chi $ 2-and F-distribution and their applications.” Biometrika, 36(1/2): 202–232.
  • Seber, G. (1963). “The non-central chi-squared and beta distributions.” Biometrika, 50(3/4): 542–544.
  • White, C. and Eisenberg, H. (1959). “ABO blood groups and cancer of the stomach.” The Yale journal of biology and medicine, 32(1): 58.

Supplemental materials

  • On the Existence of Uniformly Most Powerful Bayesian Tests With Application to Non-Central Chi-Squared Tests. Supplementary Material. The supplementary material has two parts. First part describes an R function to numerically calculate the value of the non-centrality parameter that defines the alternative hypothesis for a UMPBT. The second part provides a plot of $r(\theta)$ versus $\theta$ in Theorem 2 for several evidence threshold values $\gamma$. The plot is drawn under an example when $\theta$ is the non-centrality parameter of a $\chi^2$ distribution with 10 degrees of freedom.