by JDHURF » Thu May 16, 2013 2:27 am
The urban/suburban distinction is to me irrelevant. Clearly the suburbs surrounding the urban center enjoy the same access and use of internet utilities that the urban center does. That there may be more hate speech emanating from the suburbs surrounding the urban center than from the urban center itself seems to me what one could predict (hence my distinction between Tulsa and the surrounding suburban and rural areas previously mentioned). The important distinction is between the urban/suburban areas and the rural where, I repeat, an average 40% of the population does not even have access to the internet. Take the states of Utah, Arizona and Colorado. From the general vantage point of the heat map they appear to be among the least hateful in tweets. Is this because they are the least hateful states? To put it lightly, that is very doubtful, considering the legislated law in one of the states literally, no Godwin here, parallels Nazi policy. Zoom in, take Colorado, and I'll be damned but the heat literally encircles the labeled names of cities with the rural, barren Colorado land remaining grey. There is the anomaly of the heat circle seemingly in the middle of nowhere in between Greeley and Sterling, but this is no doubt precisely due to their "normalizing" methodology which apparently made the lone bastard with the only internet service out there who tweeted in the middle of nowhere appear as a whole population of tweeting bigots.
The map narrows the population down to people who have access to the internet, then to those who use it, then to those who use twitter and finally to those who geotagged their tweets. While there are people living in rural areas without even access to the internet, let alone twitter, the map is nothing even near a map of actual centers that emanate hate geographically, as lnrw seemed to believe, as I corrected (that being the context here).
I read their little sheep blog and I do not see anywhere within their methodology where these problems are addressed and they do not even mention the fact even though clearly they should at the least be qualifying it.
It would be analogous to comparing the hate speech of written letters among geographic areas while there is a giant disparity in literacy. Oh look, New England is more hateful in written letters because they know how to read and write while those f**k in rural Utah can't draw a star in the dirt with a stick. Wow, this is all just really enlightening, accurate, useful data. It's complete and utter nonsense is what it is.