In a former life I was into geographic information systems (GIS), spatial analysis, remote sensing, computer cartography, and satellite imagery analysis. Now, not so much, but I still dig that shit. When I was in grad school, I had multiple courses that covered these subjects including spatial statistics and advanced GIS. I’ve also always thought that stats are bullshit (i.e., you can make any point you want by massaging your data… not your wiggly bits). Anyway, I tried to figure out how to make a research topic out of spatial stats and my “stats are bullshit” hypothesis. I came up with an idea, show a positive correlation between the spatial proximity of Chinese Restaurants and Animal Hospitals/Veterinarian Clinics. I started to gather the data in a small pilot area in East Tennessee (ok, it was in fucking Knoxville) and started to massage my… er, data. I started with a 1/2 mile radius around the Animal hospitals searching for Chinese Restaurants, there was a correlation, but not strong. I extended the radius to 1 mile and boom, positive correlation. What does this prove? Not a fucking thing, but it was fun. Over the years I lost my data and test cases, also, it was in the ’90s so I was limited by the data available. I would be interested in seeing this analyis performed with proper data as it is much more abundant today…

Remember, if you hear meowing while you are in a Chinese Restaurant, get the fuck out! (This actually happened to my wife and I, but at a Japanese restaurant in Baltimore)

Share This

Trackback URI |