I wanted to see if anyone has a sample code to share on using PROC GEE to construct the marginal effect (point estimate and 95%CI in percentage points) of a binary treatment variable (1 versus 0) on an binary outcome? And how would one save the output to a sas file? This is optional: --- Also if there's a sample on achieving the same goal using multiple imputed datasets that will be great.
Answers to questions like this can often be found in the list of Frequently Asked-For Statistics (FASTats) linked from the Statistical Procedures Community page. Use the Margins macro. See Example 4 in the Results tab of the macro's documentation. The macro fits the model (using PROC GENMOD) and estimates margins and/or marginal effects.
ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.