You can’t necessarily trust geographical analysis of data if you are working from a random sample.
This entry was posted in Whatever and tagged data, datavisualization, dataviz, from-ifttt, geolocation, lies, maps, sampling, Statistics, twitter. Bookmark the permalink.