The Design of Data Visualizations

Our primary reading for this week – Edward Tufte’s Envisioning Information (1990) –  primarily concerned itself with “how to reduce the magnificent four-dimensional reality of time and three-space into little marks on paper flatlands”(119)  in an accessible and comprehensible way. It contains a wealth of examples from a variety of disciplines and vividly illustrates how seemingly subtle design choices about layout, color, and even grid lines can vastly affect the reading of a visualization. But how is this specifically applicable for digital historians? Many of Tufte’s insights focus more on the presentation of data; while he does rail against misleading “chartjunk”, he’s not particularly concerned with the methods or concerns of how the data that is then presented was originally collected and created. His usefulness really comes more into play as we consider the audience and intended (or hoped for) reach for a particular project.

Thinking back to Stanford’s Orbis Project on trade routes during the Roman Empire that we looked at during a previous week, I can’t help thinking that some of my classmates’ concerns about the accessibility of the tool might actually stem from it’s design – the interface through which we, as the audience, engage with it. The prevalent use of a strident red throughout the overlay – the header bar and the selection options in particular – fights with the surrounding blacks and dark grays for attention. This distracts the audience’s eye and makes it difficult to focus on the map at center – which contains much of the information you should actually be focusing on. Even when looking at the map itself, the blue lines work alright, but the light green starts to disappear over the water, especially when it’s placed near the more bright blue and red lines. I truly wonder if some of the information overload that occurs when looking at the project stems not from the actual complexity of using the tool, but from our eyes’ and brain’s unnecessary extra effort to parse out the information and data wanted from all those other distracting visual cues, the clutter resulting from the “1+1=3 effect,” as Tufte would label it. Though we, as historians accustomed to presenting our research as a book or essay, may tend to think of visual presentation as a surface level, aesthetic trimming, Tufte makes a compelling argument for the critical role that visualization plays in successfully conveying data and information to an audience.


I am a second-year M.A. student in Northeastern's Public History Program and received my B.A. in History from Carleton College in 2011.

Posted in Uncategorized
One comment on “The Design of Data Visualizations
  1. Bobby Samele says:

    Cool article my dude!