Hello,

I need your help with a STATA code for the logistic model with random effect.
The model analyzes the effect of individual characteristics on level of trust to police.

The formula of estimated mixed logit model looks like:

Array

where i is the number of the individual, j is the number of the country, F is the logistic function, X1, ..., Xk are the individual characteristics (such as in the ordered logit models), and Zjis GDP and CPI.
Dependent Variable = binary variable believe in police 1, 0 - otherwise.

Indep. Variables = sex, age, educhigh, educmid, marital unemployed and author uses the country–level variables GDP and CPI which include both a deterministic part that is common to all countries and a random part that reflects the specific country.


I know that the xtmelogit function is used to analyze the model.


xtmelogit police sex age educhigh educmid marital unemployed | | ....

My question: how the ending of the model should be coded with RE for GDP and CPI (Corruption Perception Index) ?

Thanks!