[Openspace] Likelihood ratio test question

[Luis Galvis]

Hi German,
The likelihood ratio test that you get under that heading is really a
diagnostic for your initial model. It tells you whether the real model you
need to estimate is the OLS, when the test shows no statistical
significance, or the alternative spatial model (i.e. LAG, ERR), when the LR
test is significant. For further reference you may want to look at page 209
of the Geoda workbook (
http://www.sal.uiuc.edu/stuff/stuff-sum/pdf/geodaworkbook.pdf).

Regards,

Luis.

On Nov 17, 2007 5:42 PM, German Muchnik Izon <german at unm.edu> wrote:
> Hi,
>
> When I run a spatial lag or spatial error model I get a
> likelihood ratio test under the heading "DIAGNOSTICS FOR
> SPATIAL DEPENDENCE". Is this a test that tells me if I was
> able to correct for spatial autocorrelation given the
> weight matrix I use to estimate the spatial models?
>
> Thank you!
> _______________________________________________
> Openspace mailing list
> Openspace at sal.uiuc.edu
> http://sal.uiuc.edu/mailman/listinfo/openspace
>