Reputation
11,494
Top tag
Next privilege 15,000 Rep.
Protect questions
Badges
1 14 44
Newest
 Nice Answer
Impact
~495k people reached

Aug
21
comment Ordered Logit regression intrepretation
The additive marginal effects bake in this interaction. Try margins r.PositionOWNPART, dydx(Frame) or margins, dydx(Frame) at(PositionOWNPART=(0 1)). @EdM's approach can be done with i.PositionOWNPART##i.Frame as the covariates.
Aug
21
comment Why the summarize command in stata always give me the wrong answer?
Take a look at xtsum for the SEs.
Aug
21
revised Why do I get “not testable” in my contrast after xtmixed in Stata?
edited body; edited title
Aug
21
revised intrepretation of ordered logit
Aug
21
revised intrepretation of ordered logit
added 81 characters in body
Aug
21
revised Why do I get “not testable” in my contrast after xtmixed in Stata?
added 16 characters in body
Aug
21
revised intrepretation of ordered logit
added 5 characters in body
Aug
21
answered intrepretation of ordered logit
Aug
6
comment fixed effects question
Have you thought about a difference-in-differences approach?
Jul
31
comment Nagelkerke value equals 1. Why?
My intuition is that the log-likelihood values are correct since that is native Stata code, which tends to be very good. You can do the calculation by hand using them.
Jul
31
revised Nagelkerke value equals 1. Why?
added 8 characters in body
Jul
31
revised Nagelkerke value equals 1. Why?
added 8 characters in body
Jul
30
revised Nagelkerke value equals 1. Why?
added 2 characters in body
Jul
30
revised Nagelkerke value equals 1. Why?
added 3 characters in body
Jul
30
revised Nagelkerke value equals 1. Why?
added 271 characters in body
Jul
30
revised Nagelkerke value equals 1. Why?
added 87 characters in body
Jul
30
revised Nagelkerke value equals 1. Why?
deleted 23 characters in body
Jul
30
comment Nagelkerke value equals 1. Why?
@Lukas I don't have any other ideas. Can you report the output of fitstat and its version in the body of the question, or at least the 2 likelihood values and the mean outcome?
Jul
30
comment Nagelkerke value equals 1. Why?
Can you add the output of the logit, the fitstat and its version (which you can get with which fitstat)?
Jul
30
revised Nagelkerke value equals 1. Why?
added 532 characters in body