I want to run a multinomial logit for panel data with fixed effects. The user written command femlogit by Dr. Knorr seems to be the best option. After trying it, I realized it does not allow factor variables as independent variable. Is there any way around this if my models need to include factor variables?
Example:
femlogit y i.x1 x2 x3, baseoutcome(0)
- factor variables and time series operators not allowed.
Thanks
Ken
Example:
femlogit y i.x1 x2 x3, baseoutcome(0)
- factor variables and time series operators not allowed.
Thanks
Ken
Comment