(1) Institute for Applied Mathematics and Statistics, University of Hohenheim, 70593 Stuttgart, Germany;(2) Department of Stochastics, University of Ulm, Helmholtzstr. 18, 89069 Ulm, Germany
Abstract:
McKeague and Sasieni A partly parametric additive risk model. Biometrika 81 (1994) 501] propose a restriction of Aalen’s additive risk model by the additional hypothesis that some of the covariates
have time-independent influence on the intensity of the observed counting process. We introduce goodness-of-fit tests for
this semiparametric Aalen model. The asymptotic distribution properties of the test statistics are derived by means of martingale
techniques. The tests can be adjusted to detect particular alternatives. As one of the most important alternatives we consider
Cox’s proportional hazards model. We present simulation studies and an application to a real data set.