Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Amin Mohseni <aminmohseni@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: Re: xtserial |
| Date | Sun, 17 Oct 2010 14:35:10 -0400 |
Dear Dr. Baum: many thanks for your help. bests amin On Sat, Oct 16, 2010 at 7:09 AM, Christopher Baum <kit.baum@bc.edu> wrote: > <> > On Oct 16, 2010, at 2:33 AM, Amin wrote: > >> I have used -xtserial- and the results suggest that there is NO first >> order autocorrelation in my error terms. (all the independent >> variables are in current time same as dependent variable) >> >> Based on this result, can I simply assume there is no need to embark >> on a dynamic panel data analysis, and a fixed effect analysis would be >> sufficient? > > I don't see the connection. The absence of autocorrelation in a panel model does not imply that its dynamics are specified properly. In writing a static model (all variables at time t), you are presuming that the full effects of the Xs are represented by the contemporaneous value of y. That is sometimes a sensible assumption, but in many contexts ruling out dynamics is a bad idea. I would try specifying this model in the DPD framework (i.e., xtabond, xtabond2) and examining the coefficient on the LDV. If it is insignificant, then the static model is justified. > > Kit > > Kit Baum | Boston College Economics & DIW Berlin | http://ideas.repec.org/e/pba1.html > An Introduction to Stata Programming | http://www.stata-press.com/books/isp.html > An Introduction to Modern Econometrics Using Stata | http://www.stata-press.com/books/imeus.html > > > * > * 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/ > * * 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/