[Openspace] Calculating regional Moran values

Julia Koschinsky koschins at uiuc.edu
Mon Nov 15 15:53:40 CST 2004


The easiest way to compute local Moran values for sub-regions within the US
is to create a new weights matrix for those sub-regions and then rerun your
LISAs and global Moran's I. If your sub-regions are based on your counties,
you can create a new sub-region file by aggregating your counties in a GIS.


Julia Koschinsky
Spatial Analysis Laboratory
University of Illinois at Urbana-Champaign

-----Original Message-----
From: openspace-bounces at agec221.agecon.uiuc.edu
[mailto:openspace-bounces at agec221.agecon.uiuc.edu]On Behalf Of Joe D.
Sent: Friday, November 05, 2004 9:39 AM
To: openspace at agec221.agecon.uiuc.edu
Cc: 'Max J. Pfeffer'; 'Zev Ross'
Subject: [Openspace] Calculating regional Moran values

We are involved in a project where it would be helpful to compare spatial
heterogeneity/autocorrelation of one region of the US to other regions.  We
seek guidance from those of you who have undertaken a similar venture.  If
the local Moran values are for each county (coterminous US) and the global
Moran is for the whole county (coterminous US), which is the best way to
calculate a Moran for a sub-US region.  Specifically, our question is this:
because the global Moran involves subtracting the global mean, X-bar, for
the variable of interest from the local values, Xi, summing the local Moran'
s yields the global Moran.  But what if you want regional Moran values?  Can
you sum the local Moran values that were calculated using the national mean,
X-bar?  Or do you need to look at it on a sub-national basis where a
 "global" Moran value is calculated for each state (or region) and then
compute the local Moran values subtracting the statewide (region wide) mean
of the variable of interest?   Thanks in advance for guidance.

Joe D. Francis
235 Warren Hall
Cornell University
Ithaca, New York 14853
Phone (607) 255-1687
Fax (607) 254-2896

More information about the Openspace mailing list