Local Moran's I vs. Gi*

Discussion created by pgibbons on Jun 6, 2011
Latest reply on Jun 14, 2011 by lrosenshein-esristaff
Hi folks,

I am conducting a hot spot analysis and am trying to figure out if Local Moran's I and Gi* can be used in the same analysis?  Both statistics indicate spatial association of neighbors.  In particular, I am stuck on whether the COType produced from the Cluster/Outlier tool (high values surrounded by other high values) are similar to the results of the HotSpot Analysis tool where a statistically significant positive Z-Score indicates spatial clustering of high values.  My analysis using these two statistics comes up with slightly different results.  Should these statistics be used separately?