Confounding usually refers to the X variables only. You can assess it using PROC REG (and many other procedures) with a fake continuous Y variable. VIF and similar measures are what you want.
@PaigeMiller, thank you very much! What I am trying is running a code with the DV and the primary IV and then running it again adding the variable I am assessing for confounding and checking for the difference in estimates output table. Is difference more than 10% then the variable is a confounder. Does that make sense or do I just look at the p value by running the code DV= IV and the variable, and if significant, consider it a confounder? Thanks!
@KKIND wrote:
@PaigeMiller, thank you very much! What I am trying is running a code with the DV and the primary IV and then running it again adding the variable I am assessing for confounding and checking for the difference in estimates output table. Is difference more than 10% then the variable is a confounder. Does that make sense or do I just look at the p value by running the code DV= IV and the variable, and if significant, consider it a confounder? Thanks!
This should have been explained in your original message. Instead you make me guess at what you want, and waste my time.
I will stick with my original answer, I feel that is a better way to address confounding than your 10% method.
@PaigeMiller, apologies for having wasted your time. Thanks again.
Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.
Register today!Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.