Geographically, I am expected the bulk of my data to come from the U.S., Canada, and Europe. In terms of longitude and latitude, this would be just below 30 degrees North Latitude continuing north to beyond 60 degrees North Latitude, and from around 15 degrees East longitude to just beyond -120 degrees West longitude. I think, quite obviously, the most data of these three regions will come from the U.S., but I will be interested to see how similar sentiments for\/against oil pipelines in Canada and growing support of indigenous rights in Canada will factor into my data set. I include Europe solely because I have seen tweets in French, Italian, and German using #standingrock, and I will be increasingly interested to see if there is any measurable drop off in #standingrock in the wake of the past week. A little background – if you were unaware, the protestors camp beside the banks of Lake Oche on the Missouri River was cleared over a two day “clean up” by North Dakota and Federal guardsmen and contractors. Obviously this was a significant setback for the opposition to the proposed pipeline, as their camp, which once housed over 1000 protestors, is no more. In my home town – Cockeysville MD, which I imagine is somewhere around 39 degrees North (lat) and -90 degrees West (long), I expect some twitter buzz, but not an extensive amount. We the people of the great state of Maryland have little dealings with either pipelines or Native American tribes. In essence, our institutional injustices center more on criminal and education injustice\/decaying infrastructure\/ and pollution of the Chesapeake Bay, than Sitting Bull or the Dakota Access Pipeline.This is not to say, however, that I do not expect a significant amount of data from people associated with the Chesapeake Bay foundation, and other groups in and around Maryland that are fighting to keep nitrate waste and other waste matter from being allowed to be disposed of in the Bay – if you want good oysters or blue crab, come soon because Maryland contains large swathes of farmland and our lovely farmers, supported by our Governor, have spent decades disposing of cattle waste, nitrate waste\/byproduct waste of other pesticides into the Bay. For reference, the water in the Bay is blue brown… and has been for some time.<\/p>\n
53\/76354 = 000695 or .07%<\/p>\n
As of now, I have a total of 76,354 tweets, 53 of whom contained the loc.# which denotes a geo-locator associated with the tweet. At .07%, the percentage of tweets with a geo-locator is less than 1 tenth of my total tweets. With respect to Yau, I find his assertion of visualization “as a medium”[1] and “a way to explore, present, and express meaning in data”[1] somewhat disconcerting with respect to Standing Rock. If one was to look at the volume of tweets concerning Standing Rock, and read the vast majority of them, you would see a substantial internet amount of relatively anonymous support. However, the relative anonymity of the Standing Rock support can be easily manipulated if you fail to consider the medium through which the data is displayed. Take for instance my location data: if our President can baselessly claim that 3-5 million (10^6th) people voted illegally, couldn’t the supporters of the DAPL claim the support for the Lakota Sioux to be much smaller citing the failure of proper corroborative data that humanizes the thousands of “people” in opposition of the pipeline. Of course upon further inspection of the data, it would be revealed that the data was taken from Twitter and less than .1% of Twitter users have a geolocation attached to their profile. Nevertheless, most people would not take that extra step to research that detail, and may be convinced by the incomplete representation of the data. When you can make claims based on nothing at all, who is to stop you from misrepresenting data to fit your narrative?<\/p>\n
Considering most of my tweets fall between 30 and 50 degrees North latitude and -100 and -120 degrees West longitude, I would imagine they are centered in and around the United States. The largest concentration of tweets stems from around 46 degrees North and -100 degrees West, which if I had to guess, would be in and around the Standing Rock reservation in South Dakota. In terms of language, all my tweets but 1 are in english, with the only outlier being “es”, which I assume is Spanish (espanol). I’d say all of this data would be reasonably expected based on the nature of #standingrock. I am, however, disappointed not to see more locations in Europe, however, if I tested each location, I am sure I would find some in Canada – keeping with my hypothesis. A fun location I noticed was -2, 140, which is in Papua New Guinea!<\/p>\n
In Boyd’s words, I feel like I have a tiny “spritzer”[2] of data from a massive data set. I supposed this relates to what she and Crawford discuss regarding “Bigger data are not always better data.”[2] With respect to my collection of over 76,000 tweets, I feel as though that would constitute big data. However, because my data comes from twitter, the resulting reality of scraping my data reveals less than 1 tenth of my data contains geo-locator information – of my 76,000 collected tweets, only 53 of them can help in mapping my data geographically with respect to longitude and latitude.. Thus, because certain facets of the data are unequal or inconsistent to the size of the data, I don’t feel as though it constitutes big data. Such facts aptly illustrate Boyd and Crawford’s assertion that “Bigger data is not always better data.”[2] I have not coordinated with my fellow environmental data farmers, but I imagine their data mimics mine, or even reveals less geographical information considering the juxtaposition between how large my whole dataset is and how small my geographically mappable data is.<\/p>\n\n\n