Hi guys

I have been trying to get the odds ratios for my logistic multilevel regression with a fe estimator.

I type my regression and then the post estimation command
xtlogit, or

But, the odds ratio I get for my primary X is way too big. It's 90,0. The variable do explain a lot of my variance in Y, but it's still too big. Does anybody know if I should use a different command or if something is wrong with my X? It's an index constructed of 7 variables measuring attitudes towards immigrants. Reducing the index to 4 variables will decrease the odds ratio to 35.

/Maria