Random effects GLS model diagnostic tests

[Peter Uttley]

Hi,

I'm new to using Stata, but I've run my model and after the Hausman and Breusch Pagan tests have concluded that random effects is best to use. I've also used a Harris-Tzavalis test and found no evidence of a unit root. The output of my model is below



I wanted to know if there are any other diagnostics tests that should be carried out to test for issues such as heteroscedasticity, autocorrelation, etc.? As far as I can understand GLS is used to correct for these possible problems. Also if there are any obvious issues with the regression output please could you point them out?

Thanks