I look at this question in a different angle. My understanding is: 1. If treat tumor_grade as a numerical variable, you assume the hazard ratio is invariant between any two adjacent levels of the tumor grade (assuming invariant covariate patterns of other risks); 2. If you treat the tumor_grade as factor, you don't assume the constant hazard ratio between the levels (assuming invariant covariate patterns of other risks).
I don't see the difference between continuous risk and discrete ordinal risk in this case. If the tumor grade is the response, there will be a difference between variables types (ordinal, continuous and nominal) since that information will be used to select appropriate models. -- View this message in context: http://r.789695.n4.nabble.com/coxph-and-ordinal-variables-tp2532149p2532299.html Sent from the R help mailing list archive at Nabble.com. ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.