Stratified Cox model and PH assumption I am learning the Cox model ad I have bumped into several articles where the models are stratified.
I've learned previously that if the PH assumption does not hold for a specific variable; then the Cox model can be stratified by that variable to make sure the PH assumptions hold.
I've asked a professor and he told me sometimes the stratified model functions equally as the fixed effects in panel data. So I can add the stratum into the Cox model even if the PH assumptions hold? And if the model is stratified, it does not necessarily mean the original Cox model is non-confirming to the PH assumptions?
 A: Stratification makes sense when you have groups whose baseline hazards are expected to be different. Therneau and Grambsch note a common situation where this makes sense (pages 44-45):

Analysis of multicenter clinical trials frequently uses stratification. Because of varying patient populations and referral patterns, the different clinical centers in the trial are likely to have different baseline survival curves ... Strata play a role similar to blocks in randomized block designs analyzed by two-way analysis of variance.

They continue to describe the tradeoffs involved:

The major advantage of stratification is that it gives the most general adjustment for a confounding variable. Disadvantages are that no direct estimate of the importance of the strata effect is produced (no p-value), and that the precision of estimated coefficients and the power of hypothesis tests may be diminished if there are a large number of strata.

So a stratified Cox model does not necessarily mean that there was a failure of proportional hazards. The investigators might just have decided that strata provided the best way to control for a covariate that wasn't of primary interest.
