Abstract
The checking and assessment of generalized linear models is a largely neglected area of many statistical investigations. Statistical packages have been developed to facilitate the routine fitting of generalized linear models but most provide few facilities to enable the assessment of the adequacy of these models. Both the informal and formal techniques of model checking can be of equal importance and utility in this respect. Six types of assessment are needed to decide on the suitability of a particular model specification: assessment of the scale of independent variables; assessment of link function adequacy; assessment of variance function adequacy; investigation of systematic departure from assumed model; outlier investigation and investigation of omitted predictor variables. Techniques are discussed, together with details of their routine implementation within a knowledge-based front-end system GLIMPSE. Wherever possible, graphical displays are used; complemented by formal test statistics as appropriate.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Aitkin, M., Modelling variance heterogeneity in Normal regression using GLIM, Appl. Statist., vol. 36 (1987), pp. 332–339.
Aranda-Ordaz, F.J., On two families of transformations to additivity for binary response data, Biometrika, 68 (1981), pp. 357–363.
Box, G.E.P. and Cox, D.R., An analysisof transformations (with Discussion), J.R. Statist. Soc., ser. B, vol. 26 (1964), pp. 211–252.
Box, G.E.P. and Tidwell, P.W., Transformation of the independent variables, Techno- metrics, vol. 4 (1962), pp. 531–550.
Cleveland, W.S., Robust locally weighted regression and smoothing scatterplots, J. Amer. Statist. Assoc., vol. 74 (1979), pp. 829–836.
Cook, R.D. and Weisberg, S., Residuals and influence in regression, Chapman and Hall: New York (1982).
Davidian, M. and Carroll, R.J., Variance function estimation, J. Amer. Statist. Assoc., vol. 82 (1987), pp. 1079–1091.
Fowlkes, E.B., Some diagnostics for binary logistic regression via smoothing, Biometrika, vol. 74 (1987), pp. 503–515.
Hammond, P. and Sergot, M.J., APES: Augmented Prolog for Expert Systems, Reference Manual. Logic Based Systems: Surrey (1984).
Hand, D.J., Statistical expert systems research, The Professional Statistician, vol. 6, issue 10 (1987), pp. 10–12.
Hinkley, D., Transformation diagnostics for linear models, Biometrika, vol. 72 (1985), pp. 487–496.
Hoaglin, D.C. and Welsch, R.E., The hat matrix in regression and ANOVA, The American Statistician, vol. 32 (1978), pp. 17–22; (Corrigenda, vol. 32, p. 146 ).
Landwehr, J.M., Pregibon, D. and Showmaker, A.C., Graphical methods for assessing logistic regression models (with Comment), J. Aer. Statist. Assoc., vol. 79 (1984), pp. 61–83.
Lee, A.H., Assessing partial influence in generalized linear models, Biometrics, vol. 44 (1988), pp. 71–77.
McCullagh, P. and Nelder, J.A., Generalized Linear Models, Chapman and Hall. London (1983).
Nelder, J.A. and Wedderburn, R.W.M., Generalized linear models, J.R. Statist. Soc., Ser. A, vol. 135 (1972), pp. 370–384.
O’Brien, C.M., One view of the GLIM/IKBS initiative (GLIMPSE) GLIM Newsletter No. 12 (1986), pp. 5–13.
O’Brien, C.M., GLIMPSE, The Professional Statistician, vol. 6, issue 10 (1987), pp. 25, 26, 33.
O’Brien, C.M., Model checking features of the GLIMPSE system, Unpublished manuscript (1987).
O’Brien, C.M., Letters to the editor, “Discussion of Mallows (1986)”, Technometrics, vol. 30 (1988), pp. 135–136.
O’Brien, C.M., The GLIMPSE System Compendium, To be published (1990).
Payne, C.D., The GLIM System, Release 3.77, Generalised Linear Interactive Modelling Manual (editor) (1985); Numerical Algorithms Group Ltd.: Oxford.
Pregibon, D., Goodness of link tests for generalized linear models, Appl. Statist., vol. 29 (1980), pp. 15–24.
Pregibon, D., Logistic regression diagnostics, Ann. Statist., vol. 9 (1981), pp. 705–724.
Pregibon, D., Score tests in GLIM with applications, In GLIM 82: Proceedings of the International Conference on Generalized Linear Models (Ser. 14, Lecture Notes in Statistics), ed. R. Gilchrist. New York: Springer-Verlag (1982), pp. 87–97.
Prentice, R.L., A generalization of the probit and logit methods for dose response curves, Biometrics, vol. 32 (1976), pp. 761–768.
Smith, D.M., Single parameter transformations and GLIM, GLIM Newsletter No. 10 (1985), pp. 20–35.
Wang, P.C., Adding a variable in generalized linear models, Technometrics, vol. 27 (1985), pp. 273–276.
Wang, P.C., Residual plots for detecting nonlinearity in generalized linear models, Technometrics, vol. 29 (1987), pp. 435–438.
Wolstenholme, D.E. and O’Brien C.M., GLIMPSE - a statistical adventure, Proceedings of the 10th International Joint Conference on Artificial Intelligence (IJCAI 87, Milan, 23–28 August 1987), vol. 1 (1987), pp. 596–599.
Wolstenholme, D.E., O’Brien, C.M. and Nelder, J.A., GLIMPSE: a knowledge-based front end for statistical analysis, Knowledge-Based Systems, vol. 1 (1988), pp. 173–178.
The GLIMPSE System Manual. ISBN 1-85206-048-4, The Numerical Algorithms Group Limited, Oxford, 1989.
Working with GLIMPSE. ISBN 1-85206-049-2, The Numerical Algorithms Group Limited, Oxford, 1989.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1991 Springer-Verlag New York, Inc.
About this paper
Cite this paper
O’Brien, C.M. (1991). Glimpse: An Assessor of GLM Misspecification. In: Directions in Robust Statistics and Diagnostics. The IMA Volumes in Mathematics and its Applications, vol 34. Springer, New York, NY. https://doi.org/10.1007/978-1-4612-4444-8_7
Download citation
DOI: https://doi.org/10.1007/978-1-4612-4444-8_7
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4612-8772-8
Online ISBN: 978-1-4612-4444-8
eBook Packages: Springer Book Archive