Hello,

Suppose I have a panel data, with T= 31 days and the number of individuals is 22. Hence, T>N. I use country-specific variables that do not change across time with yearly frequency (for the most recent year), and some factors that change daily. What model is the most appropriate against this backdrop? I was thinking about Panel-Corrected Standard Errors (-xtpcse-) or Feasible Generalized Least Squares (-xtgls-). I haven't worked with regressions for such short periods (i.e., daily).
Thanks a lot.