Summary
The problem of making inferences about the ratio of two normal populations is usually known as the Fieller-Creasy problem, and it gave rise to a controversy among fiducialists and confidence-intervalists. A Bayesian solution to such a problem when the two normal populations have the same unknown variance was presented by Bernardo (1977) using reference “non-informative” prior distributions. The solution to the case in which the variances are not assumed equal is obtained here. Some numerical results for artificial populations are given.
Resumen
El problema de hacer inferencias sobre el cociente de las medias de dos poblaciones normales, conocido como problema de Fieller-Creasy, es de interés particular en las ciencias experimentales que continuamente necesitan hacer comparaciones relativas de diferentes métodos. Desde un punto de vista Bayesiano, el problema se reduce a calcular la distribución final de dicho cociente. En este trabajo se determina la distribución final de referencia, esto es utilizando tan sólo la información proporcionada por los datos, para el caso de dos poblaciones independientes con varianzas desconocidas y no necesariamente iguales, comprobando el funcionamiento de la solución obtenida frente a los datos históricos de Cushny-Peebles y a muestras simuladas.
Similar content being viewed by others
Referencias
BERNARDO, J.M. (1977). Inferences about the ratio of normal means: A Bayesian approach to the Fieller-Creasy problem.Recent Developments in Statistics. (Barra et al eds.) 345–350. Amsterdam: North Holland.
BERNARDO, J.M. (1979). Reference posterior distributions for Bayesian inference.J. Roy. Statist. Soc. B, 41, 113–147.
BOX, G.E.P. & TIAO, G.C. (1973).Bayesian Inference in Statistical Analysis. Reading, Mass: Addison-Wesley.
CREASY, M.A. (1954). Limite for the ratio of means.J. Roy. Statist. Soc. B, 16. 186–194.
DAYAL, H.H. & DICKEY, J.M. (1977). Numerical evaluation of integrals involving products of t-densities.J. Statist. Comput. Simul., 6; 19–27.
DICKEY, J.M. (1967). Expansions of t-densities and related complete integrals.Annals of Mathematical Statistics, 38, 2.
FIELLER, E.C. (1954). Some problems in interval estimation.J. Roy. Statist. Soc. B., 16: 175–185.
GRAYBILL, F.A. (1961).An introduction to linear statistical models. New York: Mc Graw-Hill.
JAMES, A.T., WILKINSON, G.N. & VENABLES, W.N. (1974). Interval estimates for a ratio of means.Sankhya A, 36, 177–183.
JOHNSON, R.A. (1970). Asymptotic expansions associated with posterior distributions.Ann. Math. Statist., 41, 851–854.
KAPPENMAN, R.F., GEISSER, S. & ANTLE, C.E. (1970). Bayesian and fidutial solutions for the Fieller-Creasy problem.Sankhya, B, 32, 331–340.
LINDLEY, D.V. (1956). On a measure of the information provided by an experiment.Ann. Math. Statist., 27, 986–1005.
PATIL, V.H. (1965). Approximation to the Behrens-Fisher distributions.Biometrika, 52, 267–271.
RAO, C.R. (1973).Linear Statistical Inference and its Applications. 2a ed. New York: Wiley.
REY PASTOR, J., PI CALLEJA, P. y TREJO, C.A. (1969).Análisis matemático. 8a. ed. Buenos Aires: Kapelusz.
STEIN, C.M. (1959). An example of wide discrepancy between confidence intervals and fiducials intervals.Ann. Math. Statist. 30, 877–881.
TIAO, G.C. & ZELLNER, A. (1964). Bayes’ theorem and the use of prior knowledge in regresion analysis.Biometrika, 51, 219–230.
WALKER, A.M. (1969). On the asymptotic behaviour of posterior distributions.J. Roy. Statist. Soc. Ser. B, 31, 80–88.
WILKINSON G.N. (1977). On resolving the controversy in statistical inference.J. Roy. Statist. Soc. Ser. B, 39, 119–171.
ZELLNER, A. (1971).An introduction to Bayesian inference in econometrics. New York: Wiley.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Sendra, M. Distribucion final de referencia para el problema de Fieller-Creasy. Trabajos de Estadistica y de Investigacion Operativa 33, 55–72 (1982). https://doi.org/10.1007/BF02888434
Issue Date:
DOI: https://doi.org/10.1007/BF02888434
Key Words
- Bayesian Inference
- Fiducial Argument
- Fisher’s Information Matrix
- Information Theory
- Non-Informative Priors
- Ratio of two Means
Palabras Clave
- Cociente de Medias
- Distribuciones no Informativas
- Distribución Poli-t
- Inferencia Bayesiana
- Inferencia Fiducial
- Matriz de Información de Fisher
- Teoría de la Información