st: How to interprete Wooldridge test for autocorrelation in panel data

[Muhammad Anees <anees@aneconomist.com>]

Dear Statalist!

I have analysed my panel data and obtained results for Wooldridge test
for autocorrelation in panel data. I need to interprete the xtserial
test results and need further suggestions. What I think is the test
result suggesting that I need not correct for correlation and proceed
to use convential FE and RE analysis using the data? or I need the
analysis for serial correlation before analysing the data? Please
could you provide a little suggestion on estimation intentions above!
The output is as with eliminating the detailed results.


Linear regression                                      Number of obs =   16122
                                                       F( 20,  4639) =    2.48
                                                       Prob > F      =  0.0003
                                                       R-squared     =  0.0031
                                                       Root MSE      =  5.6849

                                 (Std. Err. adjusted for 4640 clusters in pid)
-------------------------------------------------------------------------------------------
Omitted Resultd
-------------------------------------------------------------------------------------------
Wooldridge test for autocorrelation in panel data
H0: no first-order autocorrelation
    F(  1,    3226) =     38.821
           Prob > F =      0.0000
*
*   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/