Tufte’s study of data density relates to our initial studies of space and composition as design students. First attempts usually illustrate, even with prior training in typography, our natural inclination to “fill” any given space with unnecessary ink. An example of poor execution of graphic density can often be seen in newspaper ads. These evenly distributed displays of stretched text and mismatched graphics lack fine detail and clear organization, all in an attempt to profit by using up the maximum amount of space. Again, Tufte’s message is to use the least amount of ink to guide the viewer straight to the most important information.
It is interesting though when he remarks “Perhaps some day statistical graphics will perform as successfully as maps in carrying information.” I would think that a graphic’s success cannot be measured entirely by it’s density, that even allowing for some ample resting space on the borders, for example, might enhance the overall display.
In his similar proposal of maximizing graphics with multifunctional graphical elements, Tufte refers to Leonard Ayre’s representation of French army in Germany during WWI. What slowed my immediate comprehension of this piece was that I could not see the exact number of divisions let alone people within them, differing from, for example, Minard’s Russian campaign graphic. I think this reminds us that much forethought is required before we organize the data, keeping in mind that which is most important and perhaps sacrificing or rather saving other thoughts for another representation. Whereas Minard’s goal is perhaps to show the humanity of war, Ayre’s graphic seems purely quantitative in nature. This example also reminds me of the importance of titling. In Ayre’s example, information titling such as “The Duration of 92 French Divisions in Germany in WWI” could easily better orient the viewer.
Space and Multifunctional Graphics…
The URI to TrackBack this entry is: http://mlkreative.wordpress.com/2007/09/18/space-and-multifunctional-graphics/trackback/
You’re right. Data density is only one consideration in the overall effectiveness of a visualization. Low density does not necessarily equal a poor visualization. However, in the examples we have seen, the higher density visualizations tend to be much more interesting.