Spurious (Spatial) Correlations


Jon •

A picture of Maine with spatial data overlaid

I was a bit out of place in my Intro to GIS course during college. Our professor gave a term assignment where we were to choose a spatial data set and perform some kind of analysis upon it. The problem I ran into was that I had no spatial data on hand, and no practical application to simulate.

Finding data to work with wasn’t a big deal. It turns out that there is an absolute wealth of information online that is both spatial and free. The problem of what to do with the data remained, though. It was at this point that I found Tyler Vigen’s Spurious Correlations, a website that draws correlations between (hopefully) unrelated data sets, such as the divorce rate in Maine and margarine sales per capita over time. I thought it would be funny to do the same with spatial data for Maine.

These are my results.

Eagles and K-Cups

With the exception of Washington County, eagles seem to make their homes in areas where household K-Cup ownership is at or higher than the national average.

Schools and Aquifers

589 of Maine’s 795 schools are within 3 km of an aquifer. This means that if you are at a school, there is a 74% chance that you are within walking distance of a significant source of groundwater.

U.S. Cellular Coverage and Oil Spill Hotspots

US Cellular offers great coverage in areas where hazardous oil spills occur the most.

Cemeteries and Broadband Coverage

3875 of Maine’s 3929 cemeteries are located exactly within broadband coverage areas (98.63%).

Red Cross Facilities and Bedrock Formation Eras

In what I can only consider to be a deliberate move, Red Cross has built their facilities in Maine on top of bedrock formed before the Devonian era. This means that if you got into a time machine and went back to the period at which the bedrock was formed beneath any Red Cross facility, you would not find sharks, ferns, trees, insects, or land vertebrates.


These datasets are in no way scientifically rigorous.