Help in interpretation of mixed model output

[NN_STAT]

Hello,

I ran a model in Stata after finding a syntax online, I wanted to ask if I ran the correct model and to ask for help in interpretation.

The study endpoint is an eye disease in animals. The animals are randomly grouped into 3 different groups (a new tested treatment vs. positive and negative controls).
For each animal, the status of the disease is tested in both eyes, making the study a repeated measures study. The response (status of disease) is ordinal.

I wanted a model that will compare the treatments vs. the response, while taking into account the obvious correlation between two eyes from the same animal (correlation exists, I checked it).

I ran this syntax I found:

xtreg  Score i.Group##i.Eye, i( Animal) mle nolog

and got the output in the attached photo.

I wanted to ask:

1. Is my model correct for a start ?
2. How do I interpret the results ? Group 1 is the new treatment, group 2 is the negative control and group 3 is the positive control.
3. My response is not really Gaussian, how badly does it effect the results ?
4. Would I get different results using xtmixed ?

An answer to ANY of these questions will be most helpful ! Thank you very much !

[NN_STAT]

In addition to my post, I ran the model again without interaction and got the attached output. The models differ significantly (p values of groups 1 vs 2).

The AIC and BIC of the first model are 206.7218 and 224.9352 respectively. The ones of the second model are attached.

Perhaps the extra information will help ?

[NN_STAT]

I also ran an ordered logistic, and the output is also attached, the code was:

ologit Score i.Group##i.Eye, cluster(Animal) nolog

I apologize for the 3 parts thread, I thought any additional information would help...