Abstract
Conditional inference has an ease of implementation that is generally unavailable with marginal inference. The main patterns for conditional inference are provided by the location and transformation families as initiated by Fisher, and by the exponential patterns as initiated by Neyman and Pearson; these are surveyed briefly together with some discussion as to how and why conditioning should be used in inference for them.
A more recent alternative pattern is provided by directional (or conical) tests and confidence methods; these lead to conditional inference for simple hypotheses with vector parameters, and can be extended to provide tests for treatment, for variance, and for treatment improvement, in the multivariate analysis of variance context.
This paper proposes the extraction of statistical frames by generalized conditioning procedures; these use versions of the conditional methods just mentioned, and then recombine the components by independence calculations. As an example, the partial likelihood of the proportional hazards model then becomes a standard likelihood for which third order asymptotics are available to give accurate tests and confidence intervals.
Similar content being viewed by others
References
Barndorff-Nielsen, O.E., Cox, D.R. (1979). Edgeworth and saddlepoint approximations,J. Royal Statist. Soc. B 41, 279–312.
Birnbaum, A. (1962). On the foundations of statistical inference,J. Amer. Statist. Assoc. 57, 269–332.
Birnbaum, A. (1972). More on concepts of statistical evidence,J. Amer. Statist. Assoc. 67, 858–886.
Brown, L.J. (1990). An ancillarity paradox which appears in multiple linear regression,Ann. Math. Statist.,18, 471–538.
Cakmak, S., Cheah, P.K., Fraser, D.A.S., Reid, N., Tapia. (1993). Third order asymptotic model: Exponential types approximation, in revision.
Cheah, P.K., Fraser, D.A.S., Reid, N. (1991). Multiparameter testing in exponential models: third order approximations from likelihood, in revision forBiometrika.
Cox, D.R. (1972). Regression models and life tables,J. Royal Statist. Soc. B 34, 187–220.
Cox, D.R. (1975). Partial likelihood,Biometrika 62, 269–276.
Cox, D.R., Reid, N. (1987). Parameter orthogonality and approximate conditional inference,J.R. Statist. Soc. B 49, 1–39.
Daniels, H.E. (1954). Saddlepoint approximations in statistics,Ann. Math. Statist. 25, 631–650.
Daniels, H.E. (1987). Tail probability approximation,Int. Statist. Rev. 55, 37–48
DiCiccio, T.J., Field, C.A., Fraser, D.A.S. (1990). Approximations of marginal tail probabilities and inference for scalar parameters,Biometrika 77, 77–96.
Evans, M.J., Fraser, D.A.S., Monette, G. (1985). Mixtures, embedding and ancillarity,Canadian J. Statist. 13, 1–6.
Evans, M.J., Fraser, D.A.S., Monette, G. (1986). On principles and arguments to likelihood,Canadian J. Statist. 14, 181–199.
Fisher, R.A. (1934). Two new properties of mathematical likelihood,Proc. Royal Soc. A144, 385–407.
Fisher, R.A. (1956).Statistical Methods and Scientific Inference, London: Oliver and Boyd.
Fraser, D.A.S. (1957). A regression analysis using the invariance method,Ann. Math. Statist. 28, 517–520.
Fraser, D.A.S. (1961a). On fiducial inference,Ann. Math. Statist. 32, 661–676.
Fraser, D.A.S. (1961b). On fiducial methods and invariance,Biometrika 48, 261–280.
Fraser, D.A.S. (1966). Structural probability and a generalization,Biometrika 53, 1–9.
Fraser, D.A.S. (1967). Data transformations and the linear model,Ann. Math. Statist. 38, 1456–1465.
Fraser, D.A.S. (1968).The Structure of Inference, New York: Wiley.
Fraser, D.A.S. (1979).Inference and Linear Models, New York: McGraw Hill.
Fraser, D.A.S. (1987). Fibre analysis and tangent models,Statistical Papers 28, 163–181.
Fraser, D.A.S. (1988). Normed likelihood as saddlepoint approximation,J. Mult. Anal. 27, 181–193.
Fraser, D.A.S. (1990). Tail probabilities from observed likelihoods,Biometrika 77, 65–76.
Fraser, D.A.S., Guttman, I., Srivastava, M. (1991). Conditional inference for treatment and error in multivariate analysis,Biometrika 78, 565–72.
Fraser, D.A.S., Massam, H. (1985). Conical tests: observed levels of significance and confidence regions,Statistical Papers 26, 1–18.
Fraser, D.A.S, Lee, H.S., Reid, N. (1990). Nonnormal linear regression; an example of significance levels in high dimensions,Biometrika 77, to appear.
Fraser, D.A.S., Reid, N. (1988). Fibre analysis and conditional inference, Proc. Second Pacific Area Statistical Conference,Statistical Theory and Data Analysis II (ed: K. Matusita) North Holland, 241–248.
Fraser, D.A.S., Reid, N. (1988). On conditional inference for a real parameter: a differential approach on the sample space,Biometrika 75, 251–264.
Fraser, D.A.S., Reid, N. (1989). Adjusting profile likelihood,Biometrika 76, 477–488.
Fraser, D.A.S., Reid, N. (1990). ¿From multiparameter likelihood to tail probability for a scalar parameter, Technical Report: University of Toronto. Revised in Fraser & Reid (1993)
Fraser, D.A.S., Reid, N. (1993). Simple asymptotic connections between density and cumulant functions leading to accurate approximations for distribution functions,Statistica Sinica, to appear.
Fraser, D.A.S., Wong, A. (1991). Approximate Studentization for marginal and conditional inference. Technical Report, University of Toronto.
Kalbfleisch, J.D., Prentice, R.L. (1973). Marginal likelihoods based on Cox's regression on life model,Biometrika 60, 267–278.
Kalbfleisch, J.D., Prentice, R.L. (1980).The Statistical Analysis of Failure Time Data, New York: Wiley.
Kass, R.E. (1989). The geometry of asymptotic inference.Statistical Science 4, 188–234.
Lugannani, R., Rice, S. (1980). Saddlepoint approximation for the distribution function of the sum of independent variables,Adv. Appl. Prob. 12, 475–490.
Neyman, J., Pearson, E.S. (1933). On the problem of the most efficient tests of statistical hypotheses,Phil. Trans. Royal Soc. A231, 289–337.
Peisakoff, M. (1951). Tranformation parameters, Ph.D. Thesis, Princeton University.
Perlman, M.D. (1969). One sided testing problems in multivariate analysis,Ann. Math. Statist. 40, 549–567.
Pike, M.C. (1966). A method of analysis of a certain class of experiments in carcinogenics,Biometrika 22, 142–161.
Skovgaard, I. (1988). Saddlepoint expansions for directional test probabilities,J.R. Statist. Soc. B 50, 269–280.
Tang, D.I., Gnecco, C., Geller, N.L. (1989). Likelihood ratio tests with nonnegative components,Biometrika 76, 577–583.
Verhagen, A.M.W. (1961). The estimation of regression and error-scale parameters when the joint distribution of the errors is any continuous form and known apart from a scale parameter,Biometrika 48, 125–132.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Fraser, D.A.S. Directional tests and statistical frames. Statistical Papers 34, 213–236 (1993). https://doi.org/10.1007/BF02925543
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02925543