论文信息 - A note on Wadley's problem with overdispersion

A note on Wadley's problem with overdispersion

SUMMARY Wadley's problem results from quantal dose-response experiments when the basic distribution is Poisson, rather than binomial. We review approaches to the problem, with particular reference to using GLIM and incorporating overdispersion. Two examples are provided, and the importance of estimating overdispersion from the fit of a saturated model is emphasized.

Byron J. T. Morgan | D. M. Smith | B. Morgan

[1] D. A. Williams,et al. Extra‐Binomial Variation in Logistic Linear Models , 1982 .

[2] A. C. Trajstman. Indices for comparing decontaminants when data come from dose-response survival and contamination experiments , 1989 .

[3] F. M. Wadley. Dosage-mortality correlation with number treated estimated from a parallel sample. , 1949, The Annals of applied biology.

[4] F. J. Anscombe. Note on a problem in probit analysis. , 1949, The Annals of applied biology.

[5] Norman E. Breslow,et al. Score Tests in Overdispersed GLM’s , 1989 .

[6] P. McCullagh. Quasi-Likelihood Functions , 1983 .

[7] Douglas G. Simpson,et al. Recursive nonparametric testing for dose-response relaflonships subject to downturns at high doses , 1986 .

[8] D. Pregibon. Goodness of Link Tests for Generalized Linear Models , 1980 .

[9] B. Morgan. Extended Models for Quantal Response Data , 1988 .

[10] N. Breslow. Extra‐Poisson Variation in Log‐Linear Models , 1984 .

[11] R. Brooks,et al. Approximate Likelihood Ratio Tests in the Analysis of Beta‐Binomial Data , 1984 .

[12] Dorothy A. Anderson. SOME MODELS FOR OVERDISPERSED BINOMIAL DATA , 1988 .

[13] Dirk F. Moore,et al. Modelling the Extraneous Variance in the Presence of Extra‐Binomial Variation , 1987 .

[14] P. McCullagh,et al. Generalized Linear Models , 1972, Predictive Analytics.