Data Visualization Activity #2: Venn Diagramming the Work of the Software Studies Initiative

Our final in-class activity on February 18th was this Venn diagram exercise–basically, an attempt to cull and visualize data from the data visualization projects of the Software Studies Initiative. (How meta!) After taking a look at the Initiative’s Projects page, we divided up the six categories (Social Media; Art and Design; Film, Video, TV, and Motion Graphics; Video Games and Virtual Worlds; Magazines, Newspapers, Books; Comics), and each member of the class chose two projects to examine–one from each of two assigned categories.

In writing this up, it strikes me that the projects of the Software Studies Initiative are organized on this page not by methodology but by the origin of its source content–this may account for the varying results of this exercise.

In any case, the “data” I asked you to cull wasn’t necessarily objective or even numeric–instead, you performed the kind of judgment-of-relevancy we found in Blatt’s article and chose “keywords.” I suggested that you look at both the topic and the methodology of your chosen projects, and the chosen keywords were quite varied as a result. Once you had your data, I asked you to visualize it using a predetermined tool–the Venn diagram. With two options, LucidChart and Visual.ly, you had options for both organization and presentation–though in hindsight I’d say that the power of LucidChart outweighs the ease of Visual.ly.

Given all of the random constraints on the “research” conducted in this activity, it shouldn’t be surprising to see that there is no consistent pattern in the results. Colby’s comparison of YouTube remixes and changes to the Google logo had very little overlap–after reading and examining her post, I’d suggest that there’s only a tiny, tiny, tiny level of connection between the two projects she examined, largely in the arena of visual creativity, and remixing or refining a single idea. It’s the sort of Venn diagram where the two circles are only slightly touching.

Laura’s two source projects both came out of Japanese media culture, which may account for the overlap in her Venn diagram–but one of these projects seems to be more about the experience of play (i.e., the literal experience of playing a video game), and the other seems to be about the results of play (the kind of linguistic play found in “scanlation”). I feel like that got a bit lost in translation in the original Venn diagram–that the keywords chosen ended up being so large that they were difficult to explain visually?

Kerishma found overlap only in the way the results of each project were presented (visualization, that is, the primary goal of the Software Studies Initiative to begin with), and saw them as otherwise being disparate, given that one analyzed a single novel and the other over 1100 feature films. Are these datasets so different? I’m not entirely certain. In any case, the difference in breadth of these sources again skewed the results somewhat, leading to keywords that were perhaps difficult to place in productive relation to one another.

When we revisit this project on the 25th, I want to tackle the following questions:

1. Did we collect the best data? If not, what data could we collect here instead?
   (One idea would be to investigate the size of the data sets for each project category, for example–or the methodologies across a single category, or the visualization choices made for each project–what variables were introduced to each data set, and why?)
2. Can we create uniform and objective rules for data collection across categories or projects?
3. What if we considered overlap not between projects, but between ideas, subjects, or methods? How would the resulting Venn diagrams differ?
4. If we can determine a meaningful dataset from the projects of the Software Studies Initiative, what would be a better way of visualizing that dataset?

Come to class next week ready to revisit this project–we’re going to try and figure out a way to improve upon our first attempt!