Hi!

I am currently conducting a panel data analysis for my master's thesis. After evaluating the F, Breusch-Pagan, and Hausman tests, I concluded to use the random effects model. I got a handout file from my Prof which now suggests xttest3 and xtserial for heteroskedasticity and serial correlation respectively, but I think after some forum posts that at least xttest3 is only possible for xtreg, fe, correct? With xtserial I did the Wooldridge test and can reject H0 - can I now just apply "xtreg, re robust" to deal with serial correlation, interpret the table and be "done"?

Many thanks in advance,
Martin