Data visualisation is the practice of transforming data into graphs, charts, maps, networks, videos, and other graphical forms. The application of data visualisation techniques can cause previously hidden patterns and trends to be identified within the data, and the resulting graphic should present the data in a more understandable and digestible format.
There are many different data visualisation techniques, and it’s important to choose a method which fits the data well and also adds meaning and truth to the data (Reas & McWilliams, 2010). The technique selected will depend on what the data is, how it’s organised, and what message the designer wishes to convey.
Data visualisation techniques can be roughly categorised based on the type of data and the interaction and distortion techniques used (Keim, 2002). Commonly used data types include one-dimensional data (such as time), two-dimensional data (such as maps), multi-dimensional data (such as data in related tables), text, hyperlinks, hierarchies and graphs. Some examples of interaction and distortion techniques include filtering, linking, projecting and zooming.
Perhaps one of the most useful features of data visualisation is its ability to allow a large amount of data to be analysed, explored, and manipulated without becoming overwhelmed by the sheer volume of information. Data is commonly explored by viewing it in a summarised form, then zooming into the areas of most interest, and then filtering out what’s irrelevant (Keim, 2002).
Data visualisation is a powerful tool when used well, however if the choice of technique does not match the data then it may fail to convey the intended message. Data which has a geographic nature is well represented by static and interactive maps, and data which has been recorded over time is effectively displayed using a time series visualisation, such as Aaron Koblin’s flight pattern animations (Koblin, n.d.). The correct application of data visualisation should result in a graphic which is easy to interpret, manipulate and analyse.
An excellent example of an interactive two-dimensional data visualisation is Nathan Yau’s Compare Worst and Best Commutes in America, which presents its data through an interactive map (Yau, 2015). Clicking on any county on the map will reveal how the average commute times across American counties compare to the chosen county. The commute time variable is split into five colour coded categories ranging from “much shorter” to “much longer”.
Nathan Yau’s interactive commute time map is a perfect example of why data visualisation provides a stronger representation of data than using tables or text. To represent the same amount of data in a table would require one column and one row per county, and it would be very difficult to compare the counties against each other without any interactive elements. Similarly, if the data was presented in text form, it would lose a lot of its meaning as it would be very difficult and tedious to read. The choice to use an interactive map to represent this set of data allows the end results to be easily and quickly understood, which makes the visualisation extremely effective.
Keim, D. (2002). Information visualization and visual data mining. IEEE Transactions on Visualization and Computer Graphics, 7(1), 100-107.
Koblin, A. (n.d.). Aaron Koblin – Flight Paths. Retrieved July 22, 2017, from http://www.aaronkoblin.com/work/flightpatterns/index.html
Reas, C., & McWilliams, C. (2010). Form + code in design, art, and architecture. New York: Princeton Architectural Press.
Yau, N. (2015). Compare worst and best commutes in America. Retrieved from http://flowingdata.com/2015/02/05/where-the-commute-is-worse-and-better-than-yours/