Geary's C Tutorial

Geary’s C is useful for discovering spatial autocorrelation.  Spatial autocorrelation means that adjacent observations of the same phenomenon are correlated in space.  Spatial autocorrelation statistics analyze the degree of dependency among observations in a geographic space and depend on a spatial weight set.  These spatial weights reflect the type of geographic relationship between polygons or points in a neighborhood.  Spatial heterogeneity describes the level of variance in the autocorrelation across geographic space.

The value of Geary’s C is between 0 and 2.  

Value of Geary's C	Meaning
C > 1	Negative spatial autocorrelation
C = 1	No spatial autocorrelation
C < 1	Positive spatial autocorrelation

 

If spatial autocorrelation is positive, the value of the variable is similar to the value in nearby locations.

Conversely, negative spatial autocorrelation indicates dissimilar variate values in nearby locations.