Statistical Vs. Infovis

In this weeks reading Gelman and Unwin 2013 put statistical graphics to the test. So it’s not the battle of the century, but an intellectual debate nonetheless. Graphic visualizations have been made popular with the public and are seen throughout the web and news, for example the New York Times. Graphics are used for qualitative and descriptive comparison of data, and report the results of data analysis (Gelman &Unwin, 2013). We have gotten used to graphs. We use graphs in excel and are taught how to do so at a pretty young age, but are we using them correctly?

Gelman and Unwin 2013 state there are two types of visual graphics, statistical and infovis. Statistical graphics tend to be traditional graphics that push to get across the point of meaningful statistical differences in order to visualize data. Infovis graphics on the other hand are more aesthetic and rely more on visual graphics to grab the audience. This is where the debate begins. Which one is better?

Unfortunately this is not an easy answer. Statistical data visualization focuses on conveying an understanding of patterns of an applied problem, and Infovis graphs focus on being attractive in order to grab one’s attention, tell a story, and look at the larger representation of a data set (Gelman & Unwin, 2013). The problem with infovis graphs is that they are usually centered around the wrong comparison of data or do not represent the data correctly. Since there is not as heavy focus on statistical understanding of a data set the wrong information is being published. Gleman and Unwin even suggest from previous studies that people have a longer retention of infovis graphics, which makes for a need of information that is published to be relevant. This is why Gleman and Unwin have made a call to action to make infovis graphics with a statistical mind set in order to push statistical graphics into the realm of infovis aesthetics creating meaningful graphics.

Reading Gelman and Unwin 2013 had made me think more about this graphical debate. According to the reading infovis graphics are meant to draw the audience toward the greater topic while statistical graphics appeal to those in the field. From this it could be seen that infovis graphics caused the inception of Gelman and Unwin’s article. Hear me out. Infovis graphs are meant to draw the audience to a greater topic, in this case statistical representation is the greater topic. Infovis graphics drew in statisticians and data analysts to the topic of data representation and has brought light to the debate of statistical vs infovis graphics. This has lead Gelman and Unwin to analyze the pros and cons of each method and suggest to combine both views. It is interesting to see how the future of visual graphics will be utilized, and hopefully the correct representation of compared data finds its way to a proper visual graphic.

Works Cited

Gelman, A., & Unwin, A. (2013). Infovis and statistical graphics: different goals, different looks. Journal of Computational and Graphical Statistics, 22(1), 2–28.