Hi,
I am looking for a standard procedure for multiple comparisons of
proportions. Basically my response variable is binary and I have two
factors. For a variable with ~ normal distribution I could use a Two-Way
Anova Fixed-Effects model with interaction. However, with a dependent
variable that has a Bernoulli distribution, what type of model should I use?
Also, for SAS users PROC GLM can perform Two-Way Anova w/ interaction. Can
I use also this procedure for what I am trying to test?
Thanks for any responses.
Regards,
LG.
|