Compare two point datasets

[HeatherCarlos]

Hi,
I have two point datasets of tobacco retailers - the gold standard is based on actual tobacco license files, the other based on the types of stores that sell tobacco (conv store, grocery store etc). 

I want to compare the two datasets - spatially, to understand how different they are.  Neither dataset has attributes in common (one has the name, the other the type) and I don't really care about the attributes.  Essentially, what I want to know: is the dataset with store type (which I can get nationally) a good proxy for the dataset based on license files (which isn't available nationally, but I can get it for a few states to test).

Any suggestions on comparing these datasets?  Anything I've come across is really getting at a relationship between their attributes rather than their location.

Thanks,
Heather Carlos
Dartmouth College

[EricKrause]

In the areas where you have license information, do the points line up with the store type points?  Can we assume the license information is complete in the areas that you have them, or are there still locations that sell tobacco legally that aren't in the license data?

You can find the proportion of license points that correspond to store locations.  You can also find the proportion of stores that are not in the licensed dataset.  If the first proportion is very high, and the second proportion is very low, that's a good indicator that store type can be used as a proxy for licensed tobacco sellers.