Spatial Statistics

2775
1
03-08-2013 04:21 AM
Highlighted
New Contributor
I am currently trying to use different spatial statistics, including Hot Spot Analysis, OLS regression and Geographically Weighted Regression. Within my data I had 'yes' and 'no' answers which was not excepted within these tools. So I changed 'no' to 0 and 'yes' to 1. I don't think this is appropriate as my results of the analysis do not match up to what is expected. Does anyone have any suggestions of what I can put instead of yes and no so I can use the data within the analysis?

And does anyone know how to merge the data output from hotspot analysis with my original data so that all of the fields are within either one table or file so I can use it all within OLS and GWR?

Any help would be appreciated!

Thanks
Reply
0 Kudos
1 Reply
Highlighted
Occasional Contributor III
Nothing! Your data is binary and all of these methods expect continuous input. You really need to do some research before throwing your data at methods. For example, with your data OLS, which is not a spatial statistical method, is regressing two data points. Since the intent of a linear model is to fit a linear regression line to the data you will always get a perfect model that is completely meaningless. 

It is time for you to talk to somebody versed in statistics.
Reply
0 Kudos