Hi everyone,
I have a panel dataset (an unbalanced one) and I have a question about robust standard errors. The code I've used is the following:
xtset country_num year
xtreg salience_mean year, vce(robust)
Essentially, I have data for different countries for a period of roughly 30 years and I want to see how the salience_mean variable grows over time.
I read that the vce(robust) option provides White's heteroskedasticity estimator. However, I'd like to do the following:
1) test for autocorrelation/serial correlation
2) use a standard error that is robust to autocorrelation/serial correlation as well as heteroskedasticity. I think this would be what some call "White-Arellano estimator" (correct me if I'm wrong)
I tried to look for an answer but it's still unclear to me how to implement the two points above.
Thank in advance for your help!