Abstract
Alternative Bayes factors are families of methods used for hypothesis testing and model selection when sensitivity to priors is a concern and also when prior information is weak or lacking. This paper deals with two related problems that arise in the practical use of these model choice criteria: sample size determination and evaluation of discriminatory power. We propose a pre-experimental approach to cope with both these issues. Specifically, extending the evidential approach of Royall (J Am Stat Assoc 95(451):760–780, 2000) and following De Santis (J Stat Plan Inference 124(1):121–144, 2004), we propose a criterion for sample size choice based on the predictive probability of observing decisive and correct evidence. The basic idea is to select the minimal sample size that guarantees a sufficiently high pre-experimental probability that an alternative Bayes factor provides strong evidence in favor of the true hypothesis. It is also argued that a predictive analysis is a natural approach to the measurement of discriminatory power of alternative Bayes factors. The necessity of measuring discrimination ability depends on the fact that alternative Bayes factors are, in general, less sensitive to prior specifications than ordinary Bayes factors and that this gain in robustness corresponds to a reduced discriminative power. Finally, implementation of the predictive approach with improper priors is discussed and possible strategies are proposed.
Similar content being viewed by others
References
Adcock CJ (1997) Sample size determination: a review. J Roy Stat Soc Ser D Stat 46:261–283
Bayarri MJ, Mayoral AM (2002) Bayesian analysis and design for comparison of effect-sizes. J Stat Plan Inference 103:225–243
Berger JO, Boukai B, Wang Y (1997) Unified frequentist and Bayesian testing of a precise hypothesis. Stat Sci 12(3):133–160
Berger JO, Mortera J (1999) Default Bayes factors for one-sided hypothesis testing. J Am Stat Assoc 94:542–554
Berger JO, Pericchi LR (1996) The intrinsic Bayes factor for model selection and prediction. J Am Stat Assoc 91:109–122
Berger JO, Pericchi LR (2001) Objective Bayesian methods for model selection: introduction and comparison. In: Lahiri P (ed) Model selection. IMS lecture notes—monograph series, vol 38. Institute of Mathematical Statistics, Hayward (with discussion)
Chaloner K, Verdinelli I (1995) Bayesian experimental design: a review. Stat Sci 10(3):273–304
Conigliani C, O’Hagan A (2000) Sensitivity measures of the fractional Bayes factor. Can J Stat 28:343–352
De Santis F (2002) Interactive use of default and robust Bayes testing methods in the presence of weak prior information. J Roy Stat Soc Ser D Stat 51(4):451–465
De Santis F (2004) Statistical evidence and sample size determination for Bayesian hypothesis testing. J Stat Plan Inference 124(1):121–144
De Santis F, Spezzaferri F (1997) Alternative Bayes factors for model selection. Can J Stat 25(4):503–515
De Santis F, Spezzaferri F (1999) Methods for robust and default Bayesian model comparison: the fractional Bayes factor approach. Int Stat Rev 67(3):267–286
De Santis F, Spezzaferri F (2001) Consistent fractional Bayes factor for nested normal linear models. J Stat Plan Inference 97(2):305–321
Dmochowski J (1996) Intrinsic priors via Kullback–Leibler geometry. In: Bernardo JM, Berger JO, Dawid AP, Smith AFM (eds) Bayesian statistics V. Oxford University Press, Oxford, pp 543–550
Efron B, Gous A (1997) Bayesian and frequentist model selection. Technical Report 193, Division of Biostatistics, Stanford University
Fisher LD, Van Belle G (1993) Biostatistics: a methodology for the health sciences. Wiley, New York
Gilks WR (1995) Discussion of O’Hagan (1995). J Roy Stat Soc Ser B 57:99–138
Iwaki K (1997) Posterior expected marginal likelihood for testing hypotheses. J Econ Asia Univ 21:1054–134
Jeffreys H (1961) Theory of probability. Oxford University Press, London
Joseph L, du Berger R, Belisle P (1997) Bayesian and mixed Bayesian/likelihood criteria for sample size determination. Stat Med 16:769–781
Kass RE, Raftery A (1995) Bayes factors. J Am Stat Assoc 90:769–781
Kass RE, Wasserman L (1995) A reference Bayesian test for nested hypotheses and its relationship with the Schwarz criterion. J Am Stat Assoc 90(431):773–795
Moreno E, Bertolino F, Racugno W (1998) An intrinsic limiting procedure for model selection and hypothesis testing. J Am Stat Assoc 93(444):1451–1460
O’Hagan A (1995) Fractional Bayes factors for model comparison. J Roy Stat Soc Ser B 57:99–138 (with discussion)
O’Hagan A (1997) Properties of intrinsic and fractional Bayes factors. Test 6(1):101–118
Royall MR (1997) Statistical evidence: a likelihood paradigm. Chapman & Hall, London
Royall MR (2000) On the probability of observing misleading statistical evidence. J Am Stat Assoc 95(451):760–780 (with discussion)
Sansò B, Pericchi LR, Moreno E (1996) On the robustness of the intrinsic Bayes factor for nested models. In: Berger JO, Betró B, Moreno E, Pericchi LR, Ruggeri F, Salinetti G, Wasserman L (eds) Bayesian robustness, IMS lecture notes—monograph series, vol 29. Institute of Mathematical Statistics, Hayward, pp 157–176 (with discussion)
Spiegelhalter DJ, Freedman LS (1986) A predictive approach to selecting the size of a clinical trial, based on subjective clinical opinion. Stat Med 5:1–13
Verdinelli I (1996) Bayesian designs of experiments for the linear model. PhD thesis, Department of Statistics, Carnegie Mellon University
Wang F, Gelfand AE (2002) A simulation-based approach to Bayesian sample size determination for performance under a given model and for separating models. Stat Sci 17(2):193–208
Weiss R (1997) Bayesian sample size calculations for hypothesis testing. J Roy Stat Soc Ser D Stat 46:185–191
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
De Santis, F. Alternative Bayes factors: Sample size determination and discriminatory power assessment. TEST 16, 504–522 (2007). https://doi.org/10.1007/s11749-006-0017-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11749-006-0017-7
Keywords
- Bayes factors
- Default Bayes factors
- Discriminatory power
- Experimental design
- Fractional Bayes factors
- Intrinsic Bayes factors
- Sample size
- Statistical evidence