Help: Interpreting ordinal logistic output

[Syeda Owais]

I'm trying to interpret a simple stata output for ologit and somehow keep getting confused:
A little about the data:
trial arm- (intervention=1, control=0)
ddepcat-(negative change=0, positive change=1, no change=2)

Q1: Just to make sure, the outputs log odds or relative risks, because I know multinomial gives you relative risks
Q2: What dependent variable group does the coeffecient for trial arm intervention show?
Q3: How do know what group was assigned to each cutpoint?

[Hong Il Yoo]

Hi Syeda, you can find an excellent answer to your questions in the Stata user guide. Please see -help ologit- and the PDF documentation therein. If you need additional learning material, you can also visit the following website: https://stats.idre.ucla.edu/stata/da...ic-regression/

[Joseph Coveney]

ddepcat-(negative change=0, positive change=1, no change=2)

In addition to looking into the resources given in Hong Il's answer above, you might want to verify that that's the ordering that you want—as opposed to something like, for example,

ddepcat-( negative change = -1, no change = 0, positive change = 1 ).