Bayesian inference is usually presented as a method for determining how scientific belief should be modified by data. Although Bayesian methodology has been one of the most active areas of statistical development in the past 20 years, medical researchers have been reluctant to embrace what they perceive as a subjective approach to data analysis. It is little understood that Bayesian methods have a data-based core, which can be used as a calculus of evidence. This core is the Bayes factor, which in its simplest form is also called a likelihood ratio. The minimum Bayes factor is objective and can be used in lieu of the P value as a measure of the evidential strength. Unlike P values, Bayes factors have a sound theoretical foundation and an interpretation that allows their use in both inference and decision making. Bayes factors show that P values greatly overstate the evidence against the null hypothesis. Most important, Bayes factors require the addition of background knowledge to be transformed into inferences—probabilities that a given conclusion is right or wrong. They make the distinction clear between experimental evidence and inferential conclusions while providing a framework in which to combine prior with current evidence.

References

  • 1. Goodman SNToward evidence-based medical statistics. 1: The P value fallacy. Ann Intern Med1999;130:995-1004. LinkGoogle Scholar
  • 2. Edwards AA History of Likelihood. International Statistical Review1974;42:9-15. CrossrefGoogle Scholar
  • 3. Fisher LDComments on Bayesian and frequentist analysis and interpretation of clinical trials. Control Clin Trials1996;17:423-34. CrossrefMedlineGoogle Scholar
  • 4. Brophy JMJoseph LPlacing trials in context using Bayesian analysis. GUSTO revisited by Reverend Bayes. JAMA1995;273:871-5. CrossrefMedlineGoogle Scholar
  • 5. Browne RHBayesian analysis and the GUSTO trial. Global Utilization of Streptokinase and Tissue Plasminogen Activator in Occluded Coronary Arteries [Letter]. JAMA1995;274:873. CrossrefMedlineGoogle Scholar
  • 6. Good IProbability and the Weighing of Evidence. New York: Charles Griffin; 1950. Google Scholar
  • 7. Cornfield JThe Bayesian outlook and its application. Biometrics1969;25:617-57. CrossrefMedlineGoogle Scholar
  • 8. Berger JOBerry DAStatistical analysis and the illusion of objectivity. American Scientist1988;76:159-65. Google Scholar
  • 9. Berry DInterim analyses in clinical trials: classical vs. Bayesian approaches. Stat Med1985;4:521-6. CrossrefMedlineGoogle Scholar
  • 10. Belanger DMoore MTannock IHow American oncologists treat breast cancer: an assessment of the influence of clinical trials. J Clin Oncol1991;9:7-16. CrossrefMedlineGoogle Scholar
  • 11. Omoigui NASilver MJRybicki LARosenthal MBerdan LGPieper Ket al Influence of a randomized clinical trial on practice by participating investigators: lessons from the Coronary Angioplasty Versus Excisional Atherectomy Trial (CAVEAT). CAVEAT I and II Investigators. J Am Coll Cardiol1998;31:265-72. CrossrefMedlineGoogle Scholar
  • 12. Goodman SNRoyall REvidence and scientific research. Am J Public Health1988;78:1568-74. CrossrefMedlineGoogle Scholar
  • 13. Royall RStatistical Evidence: A Likelihood Primer. Monographs on Statistics and Applied Probability, #71. London: Chapman and Hall; 1997. Google Scholar
  • 14. Edwards ALikelihood. Cambridge, UK: Cambridge Univ Pr; 1972. Google Scholar
  • 15. Goodman SN. Meta-analysis and evidence. Control Clin Trials. 1989; 10:188-204, 435. Google Scholar
  • 16. Efron B.Empirical Bayes methods for combining likelihoods. Journal of the American Statistical Association1996;91:538-50. CrossrefGoogle Scholar
  • 17. Hardy RJThompson SGA likelihood approach to meta-analysis with random effects. Stat Med1996;15:619-29. CrossrefMedlineGoogle Scholar
  • 18. Berger JStatistical Decision Theory and Bayesian Analysis. New York: Springer-Verlag; 1985. Google Scholar
  • 19. Edwards WLindman HSavage LBayesian statistical inference for psychological research. Psychol Rev1963;70:193-242. CrossrefGoogle Scholar
  • 20. Diamond GAForrester JSClinical trials and statistical verdicts: probable grounds for appeal. Ann Intern Med1983;98:385-94. LinkGoogle Scholar
  • 21. Lilford RBraunholtz DThe statistical basis of public policy: a paradigm shift is overdue. BMJ1996;313:603-7. CrossrefMedlineGoogle Scholar
  • 22. Peto RWhy do we need systematic overviews of randomized trials? Stat Med1987;6:233-44. CrossrefMedlineGoogle Scholar
  • 23. Pogue JYusuf SOvercoming the limitations of current meta-analysis of randomised controlled trials. Lancet1998;351:47-52. CrossrefMedlineGoogle Scholar
  • 24. Fisher R. Statistical Methods and Scientific Inference. 3d ed. New York: Macmillan; 1973. Google Scholar
  • 25. Jeffreys H. Theory of Probability. 2d ed. Oxford: Oxford Univ Pr; 1961. Google Scholar
  • 26. Kass RRaftery ABayes Factors. Journal of the American Statistical Association1995;90:773-95. CrossrefGoogle Scholar
  • 27. Cornfield JA Bayesian test of some classical hypotheses—with applications to sequential clinical trials. Journal of the American Statistical Association1966;61:577-94. Google Scholar
  • 28. Kass RGreenhouse JComments on “Investigating therapies of potentially great benefit: ECMO” (by JH Ware). Statistical Science1989;4:310-7. CrossrefGoogle Scholar
  • 29. Spiegelhalter D, Freedman L, Parmar M. Bayesian approaches to randomized trials. Journal of the Royal Statistical Society, Series A. 1994; 157:357- 87. Google Scholar
  • 30. Berger JSellke TTesting a point null hypothesis: the irreconcilability of p-values and evidence. Journal of the American Statistical Association1987;82:112-39. Google Scholar
  • 31. Bayarri M, Berger J. Quantifying surprise in the data and model verification. Proceedings of the 6th Valencia International Meeting on Bayesian Statistics, 1998. 1998:1-18. Google Scholar
  • 32. Carlin CLouis TBayes and Empirical Bayes Methods for Data Analysis. London: Chapman and Hall; 1996. Google Scholar
  • 33. Casella GBerger RReconciling Bayesian and frequentist evidence in the one-sided testing problem. Journal of the American Statistical Association1987;82:106-11. CrossrefGoogle Scholar
  • 34. Howard JThe 2 × 2 table: a discussion from a Bayesian viewpoint. Statistical Science1999;13:351-67. Google Scholar
  • 35. Cornfield JSequential trials, sequential analysis and the likelihood principle. American Statistician1966;20:18-23. Google Scholar
  • 36. Savitz DAOlshan AFMultiple comparisons and related issues in the interpretation of epidemiologic data. Am J Epidemiol1995;142:904-8. CrossrefMedlineGoogle Scholar
  • 37. Perneger TWhat's wrong with Bonferroni adjustments. BMJ1998;316:1236-8. CrossrefMedlineGoogle Scholar
  • 38. Goodman SNMultiple comparisons, explained. Am J Epidemiol1998;147:807-12. CrossrefMedlineGoogle Scholar
  • 39. Thomas DCSiemiatycki JDewar RRobins JGoldberg MArmstrong BGThe problem of multiple inference in studies designed to generate hypotheses. Am J Epidemiol1985;122:1080-95. CrossrefMedlineGoogle Scholar
  • 40. Greenland SRobins JMEmpirical-Bayes adjustments for multiple comparisons are sometimes useful. Epidemiology1991;2:244-51. CrossrefMedlineGoogle Scholar
  • 41. Rothman KJNo adjustments are needed for multiple comparisons. Epidemiology1990;11:43-6. CrossrefGoogle Scholar
  • 42. Berry DAA case for Bayesianism in clinical trials. Stat Med1993;12:1377-93. CrossrefMedlineGoogle Scholar
  • 43. Chaloner KChurch TLouis TMatts JGraphical elicitation of a prior distribution for a clinical trial. The Statistician1993;42:341-53. CrossrefGoogle Scholar
  • 44. Chaloner KElicitation of prior distributions.. In: Berry D, Stangl D, eds. Bayesian Biostatistics. New York: Marcel Dekker; 1996. Google Scholar
  • 45. Freedman LBayesian statistical methods [Editorial]. BMJ1996;313:569-70. CrossrefMedlineGoogle Scholar
  • 46. Fayers PMAshby DParmar MKTutorial in biostatistics: Bayesian data monitoring in clinical trials. Stat Med1997;16:1413-30. CrossrefMedlineGoogle Scholar
  • 47. Etzioni RDKadane JBBayesian statistical methods in public health and medicine. Annu Rev Public Health1995;16:23-41. CrossrefMedlineGoogle Scholar
  • 48. Berry DABenefits and risks of screening mammography for women in their forties: a statistical appraisal. J Natl Cancer Inst1998;90:1431-9. CrossrefMedlineGoogle Scholar
  • 49. Hughes MDReporting Bayesian analyses of clinical trials. Stat Med1993;12:1651-64. CrossrefMedlineGoogle Scholar
  • 50. Berry DA, Stangl D, eds. Bayesian Biostatistics. New York: Marcel Dekker; 1996. Google Scholar
  • 51. Berry DADecision analysis and Bayesian methods in clinical trials. Cancer Treat Res1995;75:125-54. CrossrefMedlineGoogle Scholar
  • 52. Spiegelhalter D, Thomas A, Best N, Gilks W. BUGS: Bayesian Inference Using Gibbs Sampling. Cambridge, UK: MRC Biostatistics Unit; 1998. Available at www.mrc-bsu.cam.ac.uk/bugs. Google Scholar
  • 53. Rubin DBayesianly justifiable and relevant frequency calculations for the applied statistician. Annals of Statistics1984;12:1151-72. CrossrefGoogle Scholar
  • 54. Shafer GSavage revisited. Statistical Science1986;1:463-501. CrossrefGoogle Scholar
  • 55. Walley PStatistical Reasoning with Imprecise Probabilities. London: Chapman and Hall; 1991. Google Scholar
  • 56. Tversky AKahneman DJudgment under uncertainty: heuristics and biases.. In: Slovic P, Tversky A, Kahneman D, eds. Judgment under Uncertainty: Heuristics and Biases. Cambridge: Cambridge Univ Pr; 1982:1-20. Google Scholar
  • 57. Bacon FDe Augmentis Scientarium, Book I (1605).. In: Curtis C, Greenslet F, eds. The Practical Cogitator. Boston: Houghton Mifflin; 1962. Google Scholar