Listy Biometryczne - Biometrical Letters Vol. 35(1998), No. 2, 95-106
|
THE ROLE OF COVARIATES IN EXPERIMENTAL CARCINOGENESIS*
Christos P. Kitsos Department of Statistics, Athens University of Economics and Business, 76 Patission Str., Athens 10434, Greece e-mail: kitsos@aueb.gr |
The target of this paper is to discuss the effect of covariates in the family of multistage models, which are used to describe the experimental carcinogenesis. The prediction near zero is rather difficult as any model in the neighborhood of zero can be approximated with a linear function. The class of Multistage Models was adopted to approximate the data. The effect of covariates was examined and their influence on prediction. For a real data set the logistic model was applied and the effect of the covariate "age" was examined.
percentile, prediction, covariates, response, efficiency, hazard function, link function.
