Hello,
I am working with panel Data, where t=6 and i= 550 and Stata IC16.
I have run a fixed effects regression -xtreg, fe - and afterwards regressed the residuals on the residuals of the prior period.

[predict residuals][/regress residuals l.residuals]

The result is a= 0,69 and T statistics = 47,36 which indicates strong positive serial correlation AR(1).

Question I: If I apply the Woolridge Test and use the same command, except xtserial instead of xtreg,

[xtserial math4 aexpp lunch enrollment y95 y96 y97 y98]

it indicates no serial correlation. Is this because I can not specify that I am using a fixed effects model, and is there an additional test that I can run?


Question II: What is the best option to control for Serial Correlation AR(1) ?
I have tried vce (cluster id), xtregar, and xtscc but I am really unsure which gives me the right result.

I really would appreciate some help,
thank you!

Ps. : I really try to improve the way I am asking questions by using the # command - but unfortunately it doesn't highlight my Stata code..