Abstract
This paper introduces a way of modifying the bivariate normal likelihood function. One can use the adjusted likelihood to generate valid likelihood inferences for the regression parameter of interest, even if the bivariate normal assumption is fallacious. The retained asymptotic legitimacy requires no knowledge of the true underlying joint distributions so long as their second moments exist. The extension to the multivariate situations is straightforward in theory and yet appears to be arduous computationally. Nevertheless, it is illustrated that the implementation of this seemingly sophisticated procedure is almost effortless needing only outputs from existing statistical software. The efficacy of the proposed parametric approach is demonstrated via simulations.
Similar content being viewed by others
References
Cox DR, Hinkley DV (1974) Theoretical statistics. Chapman and Hall, London
Crowder M (1995) On the use of a working correlation matrix in using generalized linear models for repeated measures. Biometrika 82: 407–410
Firth D (1987) On the efficiency of quasi-likelihood estimation. Biometrika 74: 233–245
Jensen DR (1970) The joint distribution of quadratic forms and related distributions. Aust J Stat 12: 13–22
Johnson NL, Kotz S (1972) Distributions in statistics: continuous multivariate distributions. Wiley, New York
Liang KY, Zeger SL (1986) Longitudinal data analysis with generalized linear models. Biometrika 73: 13–22
McCullagh P (1983) Quasi-likelihood functions. Ann Stat 11: 59–67
Royall RM (2000) On the probability of observing misleading statistical evidence (with discussion). J Am Stat Assoc 95: 760–780
Royall RM, Tsou TS (2003) Interpreting statistical evidence using imperfect models: robust adjusted likelihood functions. J Roy Stat Soc B 65: 391–404
Tsou TS (2005) Inferences of variance function—a parametric robust way. J Appl Stat 32: 785–796
Tsou TS (2006a) Robust Poisson regression. J Stat Plan Inference 136: 3173–3186
Tsou TS (2006b) Parametric robust test for several variances with unknown underlying distributions. Metrika 64: 333–349
Tsou TS, Chen CH (2008) Comparing means of several dependent populations of count—a parametric robust approach. Stat Med 27: 2576–2585
Wedderburn RWM (1974) Quasi-likelihood functions, generalized linear models, and the Gauss-Newton method. Biometrika 61: 439–447
White H (1982) Maximum likelihood estimation of misspecified models. Econometrica 50: 1–25
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Tsou, TS. Robust likelihood inferences about regression parameters for general bivariate continuous data. Metrika 71, 101–115 (2010). https://doi.org/10.1007/s00184-008-0204-5
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00184-008-0204-5