[Openspace] Re: Breusch Pagan

[Luc Anselin]

Breandan:
That could very well be a bug. We're looking into it.
Thanks for pointing this out.
L.


On Sep 30, 2004, at 4:37 PM, Breandan O hUallachain wrote:

>
> I am running both GEODA 0-9-5 i  and Limdep 7.0. Everything is working 
> fine and my OLS results from both programs are identical in those 
> calculations that are common to both programs. That is, R squared, F, 
> coefficients,  significance levels etc are all the same  There is one 
> curious exception.
>
> The Breusch-Pagan test for heteroskedasticity yields two very 
> different results.   Generally the value of the test statistic in 
> GEODA is a good bit larger than the result generated by LIMDEP. Any 
> ideas as to what is going on?
>
> Again this happens in OLS
>
> Breandan
>
> Breandan O hUallachain
> Professor
> Department of Geography
> Arizona State University
> Tempe, Arizona 85287-0104
> Phone: (480) 965-7533
> Fax: (480) 965-8313
> http://geography.asu.edu
>