I am performing a grouped logit regression using the command glogit on Stata 14. However, I am unable to grasp the interpretation of the coefficients. Following are some of the questions I have in mind.
1. What do the coefficients in a glogit regression represent? And how does it differ for log explanatory variables and for dummy explanatory variables?
2. How do the coefficients of a glogit command differ from the coefficients of a logit command?
3. After performing a glogit regression, is there any way to better interpret the coefficients using the margins command. I have used the command margins, eydx (varname). How do I interpret these results? Also, what would be the best option of interpret the coefficients. Should I use eydx or dydx or dyex?
4. In the results from the margins command, the Model VCE is OLS. Will this be a problem while interpreting the coefficients?
5. How does Stata compute the margins? I mean what is the mechanism behind the margins command?