Summary
The role of the inductive modelling process (IMP) seems to be of practical importance in Bayesian statistics; it is recommended that the statistician should emphasise meaningful real-life considerations rather than more formal aspects such as the axioms of coherence. It is argued that whilst axiomatics provide some motivation for the Bayesian philosophy, the real strength of Bayesianism lies in its practical advantages and in its plausible representation of real-life processes. A number of standard procedures, e.g., validation of results, choosing between different models, predictive distributions, the linear model, sufficiency, tail area behaviour of sampling distributions, and hierarchical models are reconsidered in the light of the IMP philosophy, with a variety of conclusions. For example, whilst mathematical theory and Bayesian methodology are thought to prove invaluable techniques at many local points in a statician’s IMP, a global theoretical solution might restrict the statistician’s inductive thought processes. The linear statistical model is open to improvement in a number of medical and socio-economic situations; a simple Bayesian alternative related to logistic discrimination analysis often, leads to better conclusions for the inductive modeller.
Similar content being viewed by others
References
Aitchison, J. andDunsmore, I.R. (1975).Statistical Prediction Analysis, Cambridge: University Press.
Anderson, J.A. (1975). Quadratic logistic discrimination.Biometrika,65, 39–48.
Atkinson, A. (1978). Posterior probabilities for choosing a regression model,Biometrika 65, 39–48.
Birnbaum, A. (1962). On the Foundations of Statistical Inference.J. Amer. Statist. Assoc., 269–326.
De Finetti, B. (1975).Theory of Probability 1 New York: Wiley.
Degroot, M.H. (1970).Optimal Statistical Decisions. New York: McGraw Hill.
Dickey, J.M. (1975). Bayesian alternatives to theF-test and least squares estimation in the normal linear model. InStudies in Bayesian Econometrics and Statistics (Fienberg and A. Zellner, Eds.), 515–554. Amsterdam: North Holland.
Harrison, P.J. andStevens, C.F. Bayesian Forecasting (with Discussion).J. Roy. Statist. Soc. B 38, 205–247.
Leonard, T. (1972). Bayesian Methods for Binomial Data,Biometrika,59, 581–589.
— (1973). A Bayesian Method for histograms.Biometrika,60, 297–308.
— (1977). A Bayesian approach to some multinomial estimation and pretesting problems,J. Amer. Statist. Assoc. 72, 867–874.
— (1978). Density estimation, stochastic processes and prior information (with Discussion)J. Roy. Statist. Soc. B. 40, 113–146.
Leonard, T. (1979) Why do we need significance levels?M.R.C. Tech. Report. University of Wisconsin-Madison.
Leonard, T., Low, J.A., andBroekhoven, L. (1978). Assessing the risk of fetal asphyxia.STATLAB Tech. Report. Kingston, Ontario: Queen’s University.
Leonard, T. andOrd, J.K. (1976). An investigation of theF-test as an estimation shortcut.J. Roy. Statist. Soc. B 38, 95–98.
O’Hagan, A. (1979). On outlier rejection phenomena in Bayes inference.J. Roy. Statist. Soc. B 41. 358–367.
Savage, L. (1954).The foundations of Statistics. New York: Wiley.
Schwarz, G. (1978). Estimating the dimension of a model.Annals of Statistics,6, 461–4.
Villegas, C. (1964). On qualitative probability Γ-algebras.Ann. Math. Statist. 35, 1787–1796.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Leonard, T. The roles of inductive modelling and coherence in Bayesian statistics. Trabajos de Estadistica Y de Investigacion Operativa 31, 537–555 (1980). https://doi.org/10.1007/BF02888367
Issue Date:
DOI: https://doi.org/10.1007/BF02888367