Data Visualization

Contemporary data visualization refers to "the use of computer-supported, interactive, visual representations of data to amplify cognition" (Card, Mackinlay & Schneiderman, 1999, p. 6).

Data visualizations "map data values to graphical features such as position, size, shape and color" (Heer, Bostock & Ogievetsky, 2010, p. 59) (also known as visual encodings). In doing so, data visualizations help us learn by appealing to the human visual system's ability to recognize patterns.

There are two kinds of data visualization:

• Scientific visualization uses physical data, which often have obvious spatial mappings. For example, Hurricane Katrina's clouds can be represented as just that: clouds.

Image credit: NASA/Goddard Space Flight Center Scientific Visualization Studio. More information at http://svs.gsfc.nasa.gov/3251

• Information visualization uses abstract data (economic, education, etc.), for which spatial mappings are not as obvious. Consider Charles Joseph Minard's map of Napoleon's 1812 Russia campaign. The map shows the physical location of the troops and their direction of travel, as well as temperature and the number of troops. Minard's handling of abstract data helps the user understand the devastating effects of the Russian winter on Napoleon's army.

Image credit: https://en.wikipedia.org/wiki/Charles_Joseph_Minard

Visualization provides an additional tool for exploring and analyzing datasets. Consider Anscombe's Quartet. Four datasets produce the same linear model.

Anscombe's four datasets of coordinate pairs.

Summary statistics of Anscombe's four datasets.

But when we visualize the datasets, we see just how different they are:

Image credit: https://en.wikipedia.org/wiki/Anscombe%27s_quartet

Data visualizations have yielded critical insights.  Dr. John Snow used visualization to prove that an 1854 cholera outbreak was related to contaminated water. He mapped public wells and known cholera deaths around the Soho neighborhood, and was able demonstrate a cluster of deaths around the well at Broad and Cambridge Streets. Snow’s map also exemplifies the communicative power of visualization – showing engages the audience in a way that telling doesn’t always accomplish.

Image credit: https://www1.udel.edu/johnmack/frec682/cholera/

Anscombe, F.J. (1973). Graphs in statistical analysis. American Statistician, 27, 17-21.

Card, S.K., Mackinlay, J.D., & Scheiderman, B. (1999). Readings in information visualization: Using vision to think. San Francisco, CA: Morgan Kaufmann Publishers, Inc.

Heer, J., Bostock, M., Ogievetsky, V. (2010). A tour through the visualization zoo. Communications of the ACM, 53(6), 59-67.

Tufte, E. (1983). The Visual Display of Quantitative Information. Cheshire, CT: Graphics Press.