I am doing a project on primate electrocutions. I have primate population density data - in the form of point data for each troop, and in the attribute table I have the total number of monkeys for that troop, and I have point data for electrocutions (each point is an electrocution). I have assembled these into a grid and counted number of electrocutions and number of monkeys for each square, and now I am wondering how it is best to answer my question: is electrocution density significantly associated with population density?
I was thinking of doing a Geographically Weighted Regression for this, however using the SAGA GWR tool does not seem to work - it seems to want a raster layer, or points. 'Gwr for multiple predictors' allows me to input a polygon and the columns I want for the attribute table, but it does not actually work when I try to execute it.
From a GIS and a statistical standpoint - how is it best to explore this association?
Would a correlation be better?