Re: st: RE: AR(1) Test in panel data

[Scott Merryman <scott.merryman@gmail.com>]

-xtregar, lbi- reports the Bhargava, Franzini, Narendranathan (1982)
DW statistic and the Baltagi-Wu (1999) test statistic.

The p-values are not reported but Bhargava, Franzini, Narendranathan
did generate some table of critical values (see also
http://books.google.com/books?id=uEFm6pAJZhoC&lpg=PP1&pg=PA374#v=onepage&q&f=false
):


Tables for lower and upper bonds for DW statistic

Upper and lower bounds when T = 6 (95% confidence level)

-----------------------------------------------------------------------------
                  N= 50                  N = 100               N = 150
            lower        upper      lower     upper       lower      upper
k = 1      1.8091       1.8231     1.8660    1.8731      1.8907     1.8958
k = 3      1.7954       1.8376     1.8592    1.8799      1.8859     1.8998
k = 5      1.7805       1.8517     1.8523    1.8867      1.8819     1.9046
k = 7      1.7665       1.8657     1.8444    1.8937      1.8770     1.9094
k = 9      1.7523       1.8812     1.8386    1.9018      1.8720     1.9145
k = 11     1.7380       1.8956     1.8304    1.9086      1.8671     1.9185
k = 13     1.7211       1.9118     1.8283    1.9158      1.8631     1.9233
k = 15     1.7063       1.9267     1.8163    1.9227      1.8580     1.9284

                  N= 250                 N = 500              N = 1000
            lower        upper      lower     upper       lower      upper
k = 1      1.9158       1.9183     1.9409    1.9413      1.9579     1.9586
k = 3      1.9135       1.9217     1.9393    1.9427      1.9573     1.9593
k = 5      1.9100       1.9240     1.9378    1.9441      1.9567     1.9560
k = 7      1.9070       1.9265     1.9363    1.9455      1.9551     1.9606
k = 9      1.9044       1.9296     1.9349    1.9468      1.9544     1.9614
k = 11     1.9021       1.9321     1.9333    1.9482      1.9538     1.9621
k = 13     1.8987       1.9354     1.9319    1.9497      1.9531     1.9628
k = 15     1.8956       1.9379     1.9304    1.9511      1.9524     1.9630

-----------------------------------------------------------------------------

Upper and lower bounds when T = 10 (95% confidence level)

-----------------------------------------------------------------------------
                  N= 50                  N = 100              N = 150
            lower        upper      lower     upper       lower      upper
k = 1      1.8512       1.8596     1.8953    1.8991      1.9156     1.9160
k = 3      1.8421       1.8688     1.8907    1.9037      1.9117     1.9206
k = 5      1.8338       1.8769     1.8862    1.9081      1.9076     1.9244
k = 7      1.8258       1.8851     1.8826    1.9118      1.9056     1.9245
k = 9      1.8164       1.8945     1.8780    1.9164      1.9047     1.9284
k = 11     1.8072       1.9029     1.8734    1.9209      1.9003     1.9318
k = 13     1.7999       1.9126     1.8698    1.9275      1.8971     1.9341
k = 15     1.7903       1.9209     1.8651    1.9294      1.8946     1.9375

                  N= 250                 N = 500              N = 1000
            lower        upper      lower     upper       lower      upper
k = 1      1.9336       1.9354     1.9528    1.9536      1.9668     1.9677
k = 3      1.9321       1.9370     1.9520    1.9544      1.9667     1.9679
k = 5      1.9303       1.9390     1.9509    1.9552      1.9663     1.9682
k = 7      1.9286       1.9405     1.9501    1.9561      1.9657     1.9686
k = 9      1.9270       1.9421     1.9492    1.9572      1.9652     1.9700
k = 11     1.9255       1.9445     1.9484    1.9580      1.9648     1.9705
k = 13     1.9241       1.9459     1.9476    1.9588      1.9642     1.9710
k = 15     1.9217       1.9474     1.9468    1.9569      1.9639     1.9712
-----------------------------------------------------------------------------
Where N is the number of cross sections, T is number of time periods
and k is the number of variables.

You might also find -pantest2- useful (findit pantest2 to locate and download).

Scott


Baltagi, B. H., and P. X. Wu. 1999. Unequally spaced panel data
regressions with AR(1) disturbances. Econometric
	Theory 15: 814–823.
	
Bhargava, A., L. Franzini, and W. Narendranathan. 1982. Serial
correlation and the fixed effects model. Review of
	Economic Studies 49: 533–549.


On Wed, Aug 11, 2010 at 3:23 AM, DE SOUZA Eric
<eric.de_souza@coleurope.eu> wrote:
> -findit xtserial-
>
> xtserial is a user written routine
>
>
> Eric de Souza
> College of Europe
> BE-8000 Brugge (Bruges)
> Belgium
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Hobst
> Sent: 11 August 2010 09:29
> To: statalist@hsphsun2.harvard.edu
> Subject: st: AR(1) Test in panel data
>
> Can i somehow test my panel data for AR(1)? i have seen the xtregar, but would first like to test if i really need it. and is there something similar to the durbin-watson test for paneldata?
>
> Thank you

*
*   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/