3
$\begingroup$

I am new to the realm of spatial statistics, so I need some help.

I have data that locates foreign-owned firms by XY coordinate and domestically-owned firms by XY coordinate.

I have plotted the points on a map in Arc Map.

I want to test the correlation between two maps in other words.

I want to see whether domestically-owned firms are more likely to locate where there is a foreign firm in close proximity i.e. is the location between both sets of firms not independent of one another.

I first thought of Moran's I test, but upon further reflection I am not sure it is appropriate in my case.

$\endgroup$
  • $\begingroup$ how about setting a test on the pairwise distances between the domestically-owned firms and the foreign firms using Monte-Carlo simulation? $\endgroup$ – pe-pe-rry Dec 23 '15 at 3:48
  • $\begingroup$ Okay thanks, I'm not so knowledgable on Monte-Carlo's, but I will investigate that! $\endgroup$ – Kelly Dec 23 '15 at 14:08
3
$\begingroup$

A quick answer while on Holiday: This appears to be a question in the realm of spatial point processes, and I would recommend having a look at the R package spatstat (which I'm co-developer of). We have also recently published a book which would probably be useful if you have access to that at your library. One option would be to fit a simple Poisson model for the domestic firms with the distance to nearest foreign firm as a covariate. Maybe you can find enough info to do this analysis in our free sample chapter 9, which you can download here:

http://spatstat.github.io/book/sample-chapters.html

Another approach (which may be more or less appropriate depending on the details of your data and analysis) is to use a multitype (also called multivariate) model.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.