Marginal effects for binary ologit with identical output per level

[Ghanimah Almunaifi]

Hello, I have run a binary ordinal logistic reg where my outcome variable is contributed50ormore (0 = contributed < 50, 1 = contributed 50+).

The code:

ologit contributed50ormore i.treatment i.gender i.gov i.nationality i.school i.av_h_income i.child i.level i.student_level i.married i.order

However, after running the regression then running margins, dydx(i.treatment), I got the table below.

My question is, why/how am I getting the same exact number for contributed < 50 (1) and contributed 50+ (2) but simply with a different sign? Also attached is the mtable output.

I'd appreciate any help/insights! Thanks

[Joseph Coveney]

. . . why/how am I getting the same exact number for contributed < 50 (1) and contributed 50+ (2) but simply with a different sign?

Because margins after ologit is reporting the marginal effect of the second category of the predictor (relative to the first) for both outcomes. And obviously if "Partial info" increases the proportion of the second outcome category by 5.4%, then it must perforce decrease the proportion of the remaining outcome category by the same percentage. Ditto for "Full info".

I'd appreciate any help/insights!

Use logit for an outcome with only two categories, zero and one, as you have. What prompted you to use ologit for this anyway?

By the way, what's mtable?

[Ghanimah Almunaifi]

Joseph Coveney Got it! Thank you so much, yes I used ologit by mistake. mtable just makes the margins output more reader-friendly!