Dear all,
I am running a logistic regression.
DEP: Weekly church attendace
INDEP: Job staus (4 categories) and parents church attendance when respondent was 14 (5 categories)
I am also interested in the interaction effect ONLY between Unemployed and ParentsWeekly Church attendance. I put n-1 dummies for the independent variables. What about using only the interaction term Uemployed*ParentsWeekly? If it is proper, how can I interpret the coefficient? The odd of being religious for that combination compared to all the other combinations?
thanks!
Bests,
Francesco