[Openspace] Please explain the Bivariate Moran

[koschins at uiuc.edu]

Kyle,

You can use the bivariate Moran scatterplot to relate the 
values of a variable at each location (x-axis) to the 
average values of another variable at the neighboring 
locations (lagged y). To detect statistically significant 
spatial clusters, you can run LISA maps in GeoDa.

For more detailed background information on the multivariate 
Moran scatterplot, see pp. 4-7 of Luc Anselin et al.'s 
article on "Visualizing multivariate spatial correlation 
with dynamically linked windows" at 
http://agec221.agecon.uiuc.edu/users/anselin/papers/multi_lis
a.pdf (the article refers to the old ArcView DynESDA 
extension, which is pre-GeoDa but the concepts also apply to 
GeoDa).

Julia

---- Original message ----
>Date: Tue,  6 Apr 2004 22:19:51 -0500
>From: Kyle D Peterson <KD-Peterson at wiu.edu>  
>Subject: [Openspace] Please explain the Bivariate Moran  
>To: openspace at agec221.agecon.uiuc.edu
>
>Could someone please explain to me what a bivariate moran 
will actually show?  
>This is my first time doing any spatial association and I 
am pretty ignorant in 
>the subject.  Basically I have 5 independent variables and 
one dependent 
>variable.  The dependent variable is vegetation change, and 
the others are 
>quality of life variable change.  These are in 969 meter 
cells for chicago, 
>illinois.  Could I run a bivariate moran on vegetation and 
any one of the 
>quality of life changes and get a spatial clustering 
solution?  I hope that 
>makes sense.
>
>_______________________________________________
>Openspace mailing list
>Openspace at sal.agecon.uiuc.edu
>http://sal.agecon.uiuc.edu/mailman/listinfo/openspace