Hi!
I'm working with a binary dataset with rare events (1.7% of sample). How EViews fix the small sample bias of ML method? I know Stata has a function named firthlogit.
Thanks!
Search found 1 match
- Wed Dec 05, 2018 11:38 am
- Forum: Econometric Discussions
- Topic: Penalized likelihood for logistic regression
- Replies: 0
- Views: 2765