Hello,

I would like to carry out a Cox regression analysis.

For this, I have to make sure that the variables fulfil the proportional hazard assumption. I wanted to check this with the Schönfeld residuals.

As far as I understood, I can simply save them under >Cox Regression> Save>partial residuals . Then I would plot them against the time and check if the slope is ≠ 0.

Do I have to adjust the residuals for anything else before plotting them against time. Are the residuals already risk-weighted/scaled/ adjusted for the covariance? I am a newbie to Cox regression and don't know much about these adjustments.

Thanks a lot!

Hello,

I would like to carry out a Cox regression analysis.

For this, I have to make sure that the variables fulfil the proportional hazard assumption. I wanted to check this with the Schönfeld residuals.

As far as I understood, I can simply save them under >Cox Regression> Save>partial residuals . Then I would plot them against the time and check if the slope is ≠ 0.

Do I have to adjust the residuals for anything else before plotting them against time. Are the residuals already risk-weighted/scaled/ adjusted for the covariance? I am a newbie to Cox regression and don't know much about these adjustments.

Thanks a lot!

Hi,

Thank you.

Is it possible to check the assumption for every variable one by one, entering only 1 variable a time in the Cox model and saving the partial residuals or do I have to enter all variables together in the same model, so as to adjust for the other variables, to check the assumption?

------------------------------
Viola E

Original Message:
Sent: Fri April 22, 2022 12:51 PM
From: Rick Marcantonio
Subject: proportional hazard assumption for Cox regression

Hi. From one of our statisticians:

"I don't think the user has to do more adjustments. Partial residuals are estimated for those covariates not time dependent. For every variable, they can be plotted versus time to assess the proportional hazard assumption."

------------------------------
Rick Marcantonio
Quality Assurance
IBM

Original Message:
Sent: Fri April 22, 2022 08:25 AM
From: Viola E
Subject: proportional hazard assumption for Cox regression

Hello,

I would like to carry out a Cox regression analysis.

For this, I have to make sure that the variables fulfil the proportional hazard assumption. I wanted to check this with the Schönfeld residuals.

As far as I understood, I can simply save them under >Cox Regression> Save>partial residuals . Then I would plot them against the time and check if the slope is ≠ 0.

Do I have to adjust the residuals for anything else before plotting them against time. Are the residuals already risk-weighted/scaled/ adjusted for the covariance? I am a newbie to Cox regression and don't know much about these adjustments.

Thanks a lot!