Exam reading: “Visual display of quantitative info”

For a “traditional” (i.e., not community- or critical theory-based) approach to design of graphics, I’m including two books by Edward Tufte on my reading list. There’s a certain set of information visualization people that love his work, but a set of critical theorists and rhetoricians that regards it as arhetorical and emphasizing words (and data) over visual elements. Anyway, here’s the first book on my list:

Edward Tufte. The Visual Display of Quantitative Information, 2^nd ed. Cheshire, CT: Graphics Press, 2001.

Summary: Tufte views infographics as “paragraphs about data;” however, design is universal- like mathematics rather than language. Graphics reveal data at several levels (overview to fine-grained), should be transparent, avoid distorting data, be information-dense, encourage comparisons, and be closely integrated with text, statistics, and datasets. Tufte briefly discusses the history and various types of graphics (data maps, time-series, “space-time narrative,” & relational graphics). Has chapters dedicated to graphical integrity (e.g., avoid distortions of area or scale, provide context), “data-ink” ratio=”data-ink”/total ink used to print graphic (higher better) and “chartjunk”-decorative features that don’t add interpretive info. Graphical elements should serve >1 function (e.g., position, size, color, and shape of data points can all reveal different dimensions). Also discusses data density= number of entries in data matrix/area of data graphic – idea is to try to maximize data density (maps as example of type with high data density); graphics can be smaller than we may think. Likes “small multiples”: multiple small graphics with the same format, but providing comparisons bet. changing variables (likens these to movie frames). Graphical elegance=simple design + complex data. Discusses appropriate uses of words to tables to graphics.

Comments: Outlines several reasons for continued problems with graphics: lack of quantitative training for artists, idea that statistical data are boring, idea that graphics are unsophisticated. Ideas that graphics should be transparent and conventions are universal don’t fit with rhetoric of design (though does recognize deliberate “distortion” of data as rhetorical decision). He gives several ideas for simplifying/modifying current forms; assumes people will accept these if they see them enough (resistence to new formats can be easily overcome by designers).

Links to: Kostelnick & Hassett (rhetorical design); Tversky (different emphasis on maps as data viz)