Hello,

I'm having a hard time interpreting the output of gologit2 to run an ordered logistic regression model. My dependent variable changed_freshpro is on a 5 point scale, from 1 to 5. My independent variable food_security_binary is binary where 1=food insecure and 0=food secure.

My code is

Code:
gologit2 changed_freshpro i.food_security_binary, or

Below is a portion of the output from the above code.
--------------------------------------------------------------------------------------
changed_freshpro | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]
---------------------+----------------------------------------------------------------
1 |
food_security_binary |
food insecure | .3916382 .139357 -2.63 0.008 .194985 .78662
_cons | 19.53334 5.170976 11.23 0.000 11.6263 32.81794


My question is how do I interpret this odds ratio of .3916382? My interpretation is that for food insecure individuals, the odds of being in a higher response category (2 through 5) are .39 lower than for food secure individuals.

When I look at the percentage of respondents who had a "1" for the changed_freshpro variable, this is the only interpretation that makes sense. 11.56% of food insecure individuals had a 1 for changed_freshpro, compared with 4.87% for the food secure individuals.

Many thanks,
Alyssa Beavers