A0638
Title: Goodness-of-fit test for Cox model under isotonic constraint
Authors: Huan Chen - University of Texas at Dallas (United States) [presenting]
Chuan-Fa Tang - University of Texas at Dallas (United States)
Abstract: Cox proportional hazard model has been widely used as it shows the effect of covariates on the hazard. However, in many applications researchers are only willing to assume the hazard is isotonic to covariate such that the Cox proportional hazard model may introduce biases. In this case, a general isotonic proportional hazards model is more accurate. In order to determine which model is more appropriate based on the researchers' data, a likelihood ratio goodness-of-fit test is proposed via the bootstrap method, where the null hypothesis is Cox proportional hazard model while the alternative is the isotonic proportional hazards model. Starting from the time-independent univariate cases, the test can be generalized to time-dependent univariate and partial linear multivariate cases. The pseudo-iterative convex minorant algorithm is used for the estimation of the monotone hazard to guarantee efficiency.
