ArcMap help - spatial autocorrelation variables and Moran's I

01-13-2015 01:27 PM
New Contributor

Good day,


I desperately need some guidance in using some specific tools in ArcMap.


I need to calculate spatial autocorrelation Moran's I statistics for area crimes and extract the spatial crime lags for each area in order to include it in my regression models elsewhere. I have the data in ArcMap and have joined the crime counts to the area full extent geography.


Research has suggested that calculating the distance band from the nearest neighbor and analysing outputs from the incremental spatial autocorrelation tool in order to inform generation of a spatial weights matrix file which can be used in the spatial statistic tool for calculating Moran's I is a good procedure to follow. Is this correct? It also appears to me that several of these procedures including the Moran's I tool do not appear to use the crime count itself in the calculation which concerns me because I am looking for the spatial autocorrelation between the crimes across areas rather than a measure of autocorrelation in other features. Can you shed any light on this please?


I read somewhere that the incremental spatial autocorrelation tool in version 10 does not work so well. There is a lack of information online to help with these issues and ArcGIS help does not go into sufficient detail. I wonder if anyone can also point me to a good source for guidance with ArcGIS uses and how these tools work?


Many thanks for any help you can provide.

0 Kudos
0 Replies