我也有这个疑问,我收集了一段xtgls和xtpcse的比较,但没有与固定和随机效应的联系。
xtpcse and xtgls estimate linear panel data models using feasible GLS. xtpcse computes OLS estimates with panel-corrected standard errrors, while xtgls computes feasible GLS estimates. These commands allow estimation in the presence of AR(1) autocorrelation within panels, as well as heteroskedasticity or cross-sectional correlation across panels. In the case of cross-sectional correlation, xtgls requires T > n.