The Infovis of the Beholder
In the publication “Infovis and Statistical Graphics: Different Goals, Different Looks”, Gelman and Unwin (2013) describe the differences in the graphics typically designed and presented as statistical or informational visualizations. They provide excellent descriptions of the differences in the reasoning and target audience for the two graphical presentation styles. Statistical analysis often focuses on objectivity and replication; these may be less visually appealing, even boring, because the focus is on the actual data, and analysis or data model employed. The designer looks to impress more with the analytical results then with the overall appearance of the figure. The alternative, informational visualizations, is more about appeal, or the “wow factor.” The focus here is on creatively, or drawing the readers’ attention to a particular point or element. The article defined the differences and suggested that both domains take a page for the other’s book.
I spent some time looking at the graphics this article cited and drawing some conclusions about other figures I have encountered when reading other materials, both empirically researched and consumer focused. I began to think about the graphic images that I found interesting and what drew me in. I remember one particular timeline that was simple and straight forward, but engaged my interests because it had elements that were personally meaningful and evoked deeper thought. By engaging me as the audience with elements that were meaningful I paid closer attention to the entire graph. As I perused the timeline I periodically hit upon a point of personal interest; this caused me to stop and reflect on the timeline as a whole. I translated the timeline into my own life experience, and with that it became meaningful and memorable. Statistical graphs are not typically personally meaningful. They are more likely to be subject specific and focused. What if elements could be implemented that would draw in the audience? It is not a far stretch to imagine that an audience has some initial interest due to the fact they are reading the material in the first place. Could engaging elements be designed based on the topic as a whole, that might draw an audience in more? While this method was not directly suggested by Gelman and Unwin (2013), I believe it could increase the interest/engagement of statistical graphics for readers.
My current academic career requires me to read a great deal. This reading often includes graphical information. I have been exposed to figures I felt were very interesting and informative as well as those that were confusing and off putting. Wainer (1984), gives many examples of poorly designed and executed graphical images. To juxtapose the poor examples, the publication offers an exemplary graphic; it is often included as the gold standard of graphical images. This figure (image 1), is referred to in many circles as “the best statistical graphic ever drawn” (Jacobs, n.d.).
This is the Minard Map, Tufte (2001) refers to this image in The Visual Display of Quantitative Information. The data depicted describes skirmishes between Russia and Napoleon’s army in 1812. The image includes six different elements of data; included are geography elements, the army’s course, the army’s direction, the remaining number of soldiers, temperature, and time. I admire the scope of this image but will admit I find it a bit over my head. I understand though why it is presented as exemplary. It offers elements of what Gelman and Unwin (2013) discuss as being both statistically and visually informational.
I have to close my comments with my own example of poor graphical information, (image 2).
For those of you unfamiliar with the reference to the RSS Boaty McBoatface, in 2016 there was an effort by the Natural Environment Research Council (NERC) to enlist the help of the public to name a new polar research ship (Ellis-Peterson, 2016). As may be determined from the graphic above, the winner was……”RSS Boaty McBoatface”. As exemplified in the Wainer (1984) publication, this graphic is unclear, funny, but unclear. The graph attempts to compare the number of votes of the five most popular names but gives us little else. What is the point for this last graph you ask? I wanted to find my own example of poor graphics AND I wanted to mention the RSS Boaty McBoatface as part of my graduate writing.
References
Ellis-Peterson, H. (2016, Apr). The Guardian. Boaty McBoatface wins poll to name polar research vessel. Retrieved from https://www.theguardian.com/environment/2016/apr/17/boaty-mcboatface-wins-poll-to-name-polar-research-vessel
Gelman, A., & Unwin, A. (2013). Infovis and statistical graphics: different goals, different looks. Journal of Computational and Graphical Statistics, 22(1), 2–28.
Jacobs, F. (n.d.). Big Think. The Minard map-the best statistical graphic ever drawn. Retrieved from http://bigthink.com/strange-maps/229-vital-statistics-of-a-deadly-campaign-the-minard-map
Tufte, E. (2001). The visual display of quantitative information. Graphics Press, Cheshire, USA,, 4(5), 6.
Wainer, H. (1984). How to display data badly. The American Statistician, 38(2), 137–147.
