03-05-2017 08:19 PM
I have a question about the method to see if the variables follow the proportional hazards hypothesis.
In SAS help, “The methods are derived from cumulative sums of martingale residuals over follow-up times or covariate values. ” is described. But I think that the method is following below the procedure.
1) At interesting covariate, generate random number for survival data with hazard ratio of that is obtained by the raw data
2) Calculate the Schoenfeld residual based on generated random number
3) Calculate cumulative the Schoenfeld residual by sorting the Schoenfeld residual in chronological order
4) Calculate the maximum absolute value of the cumulative Schoenfeld residual
5) Repeat 1) - 4) to generate distribution of simulated the cumulative Schoenfeld residual
6) obtain the P-value that the rareness of the raw data in the distribution
Could you point out my mistake?
Thanks in advance,