The more red an area is, the more hate the mappers found. Check out the findings:
The makers logged geo-coded tweets between June 2012 and April 2013. They filtered the tweets using a list of derogatory terms, and then manually went through the list to identify the tweets used in a negative manner. They were left with 150,000 tweets on which to base these maps.
The terms used were pulled from Wikipedia entries on racist and anti-LGBT slurs, and can be found here (I won’t repost the list because some of them are pretty offensive).
In terms of what the maps actually represent, well, it’s not exactly a pretty picture. “We think that 150,000 [tweets] is a sufficiently large number to be quite depressed about the state of bigotry in our country,” write the researchers.
In a FAQ about the maps, the researchers emphatically point out that the maps are “decidedly NOT showing population density.” Though the east coast data seems to reasonably line up with population distribution, the mappers looked at tweets on a county-by-county basis, then “normalized [the derogatory tweets] by the overall number of tweets in that county.” They point to the lack of activity in California – the most populous in the U.S. – as evidence that population density isn’t responsible for their results: “There are many possible explanations for some of the distributions that you can see, and we don't pretend to have all of the explanations. But population just isn't one.”
Lastly, the researchers make it clear there isn’t really a defense for this prevalence of derogatory language: “If you are a disgruntled white male who feels the persistence of hatred towards minority groups is a license to complain about how discrimination against you is being ignored, just stop.”