st: test for autocorrelation

[Katja Hillmann <katja.hillmann@wiso.uni-hamburg.de>]

Hi,

I have a panel with 36 cross-section over 17 time units. I'd like to 
figure out, wether the autocorrelation structure of my data set follows 
an AR(p)- or/and MA(q)-process. Is there any tool/command in STATA which 
tests for the (non-)presence of Autocorrelation?

I know, for times series I plot the AC and the PACF of the stationary 
times series in order to identify ARIMA-Modell. Or may I simply plot the 
AC and PAC of the time series as an average over all cross-sections? 
Does it really reflect the autocorrelation of the panel.

In the end I want to run a fixed-effects regression (xtregar).

Thanks a lot!

Kate

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/