Hi Statalist,
I am running a hierarchical generalized linear model. My outcome is binary (0, 1) and my data is students nested in schools, further nested in countries. I have two data samples, one from Western countries, one from Arab countries, and I am running my regressions separately for each sample. I am using the exact same outcome, independent and control variables in my models. If I would like to compare the coefficients across models (e.g. SQE), my understanding is that I would need to compare the coefficients and test whether the difference between coefficients in the two models are statistically significant using a test for equality of coefficients.
I am unable to share my data, but below is the code I am using for my models. I use the exact same regression for both samples.
I guess I am asking whether my understanding is correct here?
I am running a hierarchical generalized linear model. My outcome is binary (0, 1) and my data is students nested in schools, further nested in countries. I have two data samples, one from Western countries, one from Arab countries, and I am running my regressions separately for each sample. I am using the exact same outcome, independent and control variables in my models. If I would like to compare the coefficients across models (e.g. SQE), my understanding is that I would need to compare the coefficients and test whether the difference between coefficients in the two models are statistically significant using a test for equality of coefficients.
I am unable to share my data, but below is the code I am using for my models. I use the exact same regression for both samples.
Code:
xtmelogit admin SQE female age job education || idcountry: || idschool: , or
