Dear all,
we have a problem calculating test statistics for serial correlation in
panels.
In the DPD98 software package for GAUSS, tests for first-order (m1) and
second-order correlation (m2) are included by default. Using the
xtabond2 command in Stata also produces these two test statistics for
GMM-in-systems and GMM-in-differences.
We now want to obtain a similar test statistic for the OLS in levels
(regress) and within-groups (xtreg , fe) estimation. The Durbin-Watson
and Breusch-Godfrey (bgodfrey) tests do not work (error: "sample may not
include multiple panels". With the Wooldridge-Drukker test, we obtain a
measure only for first-order serial correlation.
Any help is highly appreciated!
Christian Andres and Andr=E9 Betzer
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
