03-27-2017 02:10 AM
I would like to compare the effects of variables between men and women to see whether the effects of the same variable on a dependent variable are different by sex. So I construct men's and women's data, respectively. In my research, a response variable is obesity(0=obese, 1=normal) and I run a logistic analysis.
I get the following results:
Men Women Men vs. Women
Education -.037(.002)*** -.047(.011)*** significant difference?
Income -.122(.003)** -.075(.032)*** significant difference?
I can say that education and income are related to obesity for both men and women. However, I also wonder whether there is a difference of the effects of education and income on obesity between men and women. This requires a comparison of two coefficients from men and women for the same variable. How can I test whether there is a difference of coefficients for each variable, education and income between men and women in SAS?
03-27-2017 03:14 AM
03-27-2017 06:54 AM - edited 03-27-2017 06:57 AM
You need combine men's and women's data into one table, and use ESTIMATE . class SEX; ........... estimate 'F vs M' sex 1 -1 ;