Dear Users,
xtregar reports the LBI test statistic for the presence of no serial
correlation. As there is no table, to my knowledge available as it will
clearly depend on the number of groups, periods, and variables, I came
up with the following "methodology" for which I would like to invite you
to comment.
I simply referred to the tables in
Bhargava, A., L. Franzini and W. Narendranathan, (1982), "Serial
Correlation and the Fixed Effects Model," Review of Economic Studies,
49: 533-549
to check if rho=0 can be rejected. If then I am safely in the area of
serial correlation I conclude that modelling the error term as an AR(1)
process is indicated.
Results are then checked estimating panel corrected standard error
(PCSE) estimates for linear cross-sectional
time-series models where the parameters are estimated by OLS or
Prais-Winsten regression, i.e. using xtpcse.
Any comments?
Iwan
_________________________________
Iwan Barankay
Department of Economics
University of Warwick
Coventry CV4 7AL
U.K.
Tel.: **44 (0)2476524930
Fax.: **44 (0)2476523032
Email: [email protected]
Web: http://www.warwick.ac.uk/~ecrhc
*
* 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/
