The Visual Display of Quantitative Information, 2nd edition [Hardcover]

Quite insightful, especially the historic perspective. The only difference between the newer and 1983 edition is color is added to a few of the charts.

Edward Tufte''s "'The Visual Display of Quantitative Information"' provides a very clear introduction to quantitative graphic design. Embedding graphic examples in the text allows the reader to immediately understand the concepts Tufte describes, even if the concepts are unfamiliar. Historical graphics (especially Playfair''s graphics) make the reading especially interesting, because they show the reader where the concepts of statistical data visualization originally come from. I also found that Tufte''s habit of including a conclusion to each chapter was very helpful, because it summarizes (often in point-form) the key points of the chapter. Clearly distinguishing important design points with indents, italics, and bullets also helped me remember them as I read through the book.

Tufte''s frequent comparisons with "'good"' and "'bad"' design allowed me to think about graphics in a more analytical way. His emphasis on simplicity and efficiency also made me consider my own work. I found that some of Tufte''s principles can be applied beyond visual representations: in writing, for example, the author can strive for greater clarity without cluttering their sentences with an elaborate writing style. (In graphic design, this would be similar to eliminating chartjunk.)

Tufte, from his introduction and his writing style, is obviously very passionate about the integrity of data graphics. He often expresses outrage over the data distortion that some designers introduce in overly-artistic graphics, and denounces the attitude artists have towards data graphics (they have "'contempt"' for the visualizations). While distorting data is a serious offence, I found that his use of language in describing inefficient graphics was itself often overly negative, sometimes to the point of being distracting. If one of Tufte''s main principles is that the design should not distract from the data itself, then perhaps he could have used a slightly more neutral writing style to in order to more clearly convey his principles.

A truly great and ground-breaking book. It was as much a pleasure to read again in 2012 as it was the first time I encountered it 25 years ago. Its analysis and advice are at least as relevant now as they were at the original publication date. (A few examples from the "computer era" are visually dated, but you can see slicker versions of the same problems and errors all over the web and on mobile device apps.)

For those who don't know the book, it is an analysis of what makes a truly great presentation of quantitative information work -- where the meaning of "truly great" includes easy to examine, easy to understand, multi-levelled (you can get information at various levels of detail just by how you look at it), accurate and informative (not misleading), visually beautiful, parsimonious, and inviting or intriguing. It invites and challenges the reader to think about how data presentation can be designed for the user to get the most out of it with the least cognitive load, misunderstanding, and distraction, and the maximum of delight and engagement.

The design and presentation of the book exemplifies its message, with diagrams and examples integrated with the text so that each reinforces the other, and a graceful and helpful use of typography and text layout to bring out the important messages. Tufte is an engaging and personal writer, and he is direct about what he likes and doesn't like and why. Because he is insightful, brilliant, and witty, his prose style too is part of the book's appeal.

An essential on the shelf of anyone concerned with communicating complex information visually, including people like me who design software user interfaces.