I have a short panel of data in our analysis. (77 countries, for over 4 years with a total of 308 observations). I already tried using xtpcse command in Stata to correct the autocorrelation and heteroskedasticity in the model; it did not work since there are no time periods common to all panels due to missing data. I am wondering if there are other tests that can help to correct the autocorrelation and heteroskedasticity in the model.