coef covariance method

[tommyez]

Hi everyone, currently I am doing fixed effect regression, but I have heteroscedasticity and autocorrelation problem, I want to use robust standard error as the treatment of my problem. My Question is, which coef covariance method should I use? Thanks

Best Regards,
Tommy

[EViews Glenn]

White period allows for general correlation structures within a cross-section. It's clustered standard errors, where the cross-section is the cluster identifier. Note that this is more general than autocorrelation.