The visual world of Aesthetics and Novelty!
Wow, that doll is so beautiful! I want it dad, could you please buy it for me. As a child we have always craved for beauty and as we grow up we look for it everywhere. Beautiful things appeal more to us than the drab ones. Even while dating people give preference to an attractive face rather than a caring heart. It has been human nature to be in awe of prettiness, while ignoring the vile stuff around us whether it is tangible or intangible. This week’s reading focused on the resplendence of data visualizations, as the ways to attract readers, all the way while criticizing them as how they fail to reveal the hidden effects in data.
Gelman and Unwin (2013) alleged that graphics are being mistreated even today, while quoting the fact that only 20% of the articles in the Journal of Computational and Graphical Statistics relate to graphics. Although, graphics provide context to the readers, arousing their interest, yet they are underestimated. As Tukey highlighted that graphs can be used for easy comparison while being persuasive, but they should be provided with a proper supportive analysis rather than just being glamorous on their own. Authors of this article laid down 6 basic rules for the data visualizations, which governs the communication of discoveries in an efficient way and then explored the two practices which have been employed for the purpose of visualization, namely Statistical data visualization and the Infographics commonly called as Infovis (Information Visualization). They presented the best data visualization projects of the year, with a constructive criticism in order to emphasize the importance of embellished visualizations for statistical graphs. Most of them were very good in capturing the attention of the readers, but they did not represent the data quite well and were difficult to interpret. For example, a CoxComb plot depicted in the reading although grabbing my attention never made sense to me, and I had to search over the internet and read about it before I could drive out some interpretation from it. This suggests that the goal for the two are entirely different, while statistical graphs intends to uncover the hidden patterns in data, Infovis are designed to capture the attention of the readers. Further, because of their very nature of goals, it is beneficial to use them together. Firstly, use the Infovis to grab the attention, absorbing them in the context and then use the vapid statistical graphs to show the patterns.
People tend to use visuals which they think are are charming. They use unwarranted 3D pie-charts without realizing the fact that people are very bad in comparing the areas. While browsing the best R-packages used for data visualizations I came across threeJS. There I found a couple of awesome visualizations worthy of admiration, but at the same time I failed to understand their purpose apart from looking beautiful. A website called Roberta di Camerino depicts an eye-catching visualizations saying that it is the story of their brand and journey. While appealing to me at first, as I scrolled down till the end, I found it to be very annoying way to convey some story. Similarly, there is a website called Codeology, depicting stunning visuals of GitHub profiles of various users. They say that it brings the code to life visually, by representing the shape and color with the coding language used and size being the length of code. Below is the visual they generated for my GitHub profile. While the visual appeared quite exquisite to me, I failed to understand the purpose of it.
Generally, in most of the cases, what makes a visual a good statistical graphic also makes it a bad Infovis and vice-versa. We must strive for a balance between the two, through the cooperation of designers and data analysts. Moreover, what is a adorned as a great Infovis today, might become a cliched statistical graph tomorrow as these Infovis serve a purpose of inspiring the statistical graphs of tomorrow.
References:-
Andrew Gelman & Antony Unwin (2013). “Infovis and Statistical Graphics: Different Goals, Different Looks”. Journal of Computational and Graphical Statistics, 22:1, 2–28, DOI:10.1080/10618600.2012.761137
The top ten worst graphs. Retrieved from https://www.biostat.wisc.edu/~kbroman/topten_worstgraphs/
