Customer Reviews

Graphic Discovery: A Trout in the Milk and Other Visual Adventures

5 star:		(6)
4 star:		(0)
3 star:		(2)
2 star:		(1)
1 star:		(0)

 

 

I ordered this book with great anticipation. I had read the previous review which calls it the best mathematical book of 2005. I was sorely disappointed. The writing is extremely wordy and many of the digressive footnotes served the author better than they serve the reader. The author is overly proud of his having known John Tukey and he also subjects us to a tediously long example involving his son's Princeton acceptance letter.

I think the author is actually a fine fellow who genuinely loves graphs and charts, and he does manage to present many classic pieces of graph design advice in a congenial way. But the essays are rather disconnected. The author is fairly good on mathematical graph design but his ventures into related issues such as collation and ordering and document design are not successful and his lack of expertise in these areas is painful.

The book ends with a very odd twenty-page biographical dictionary, which partly covers people prominent in the history of graphing, but also includes random folks who just seemed to have caught the author's attention or who were mentioned peripherally in examples in the text, such as Seneca, Henry David Thoreau, and all (I think) of the current American Supreme Court Justices.

Overall, this book is a kind of brain dump which feels like reading the backup copy of the author's future-projects file. There are other better places to start learning about graph design, including Tufte, Cleveland, or even the old standby "How to Lie With Statistics."

A delightful and thought-provoking book on statistical graphics. Wainer provides compact case studies of how graphical presentations such as bar charts, plots, and scattergrams can lead to important discoveries. The most compelling examples show how a published graphic could be dramatically improved to avoid misleading interpretations or make new discoveries. The most entertaining parts are his vignettes of historical figures, such as his twin heroes of William Playfair and John Tukey. I enjoyed Wainer's sardonic wit, personal anecdotes, and popular culture references, but the real gift was the clarity of thinking and the wise guidance about deep issues in statistics, data mining, and information visualization.

Ben Shneiderman, College Park, MD

[...]

You may think it's too early to already be picking the best math book of 2005, but I can't imagine I will regret this choice. Wainer's book is accessible to people with a minimal statistical background, with its fascinating voyage through the history of data displays. It provides interesting biographical information on some of the characters we meet along the voyage. Then it finally looks to the future direction of data analysis.
I can't imagine any mathophile, and particularly any teacher of statistics at any level, who won't find this book a treasure trove of delights. Highly recommended.

If you have read Tufte's books and, especially, W.S. Cleveland's books, you do not need to buy or read this book. You've got most of it and much more. I may recommend this book only as a first and fast approach to this matter.Then go directly to the Masters.

I've been looking for a way to jazz up my undergraduate statistics course. On a whim, I bought a copy of Wainer's "Graphic Discovery." What a treat! This is a book that should be required reading for every undergraduate in psychology. Beautifully typeset and written, loaded with interesting and important examples, the book manages to be entertaining and informative without being condescending or simplistic. Wainer has a real gift for elegant, succinct writing. Highly recommended!

James H. Steiger, Professor
Dept. of Psychology and Human Development
Vanderbilt University
Nashville, TN 37203

If you work with data and have to communicate it with others this book is a must read. There is beauty in simplicity and Wainer brings it to the fore in Chapter 5. His focus on Playfair and Tukey gives us good historical insighe. It just so happened that the first example I read was Chapter 11 'Order in the Court' and it solved two problems in a totally unrelated field for me. The moral is that alphabetical order is great for finding information but useless when you are using the information to make judgements.

This book is a true gem on par with Tufte's "The Visual Display of Quantitative Information". For statisticians, it might be even more important than Tufte's book. The book covers many different techniques for the display of data, but goes beyond that with some historical developments and excellent discussions about the use and misuse of the techniques. Simpson's paradox is one of the techniques that is discussed several times to explain how opposite interpretations can be obtained for an entire data set and for subsets of the data set. Another interesting example discusses some work by Abraham Wald, who was trying to determine where to add extra armor to planes on the basis of the pattern of bullet holes in returning aircraft. His conclusion was to map the holes on the planes that had returned and then put extra armor everyplace else. His key insight was that the planes with holes that were observed could still return to the base, while the planes that did not return must have been hit somewhere else. In all, a delightful read.

As an individual who has been removed from the field of statistics and graphical presentation for some time now, I found this book enjoyable, informative, and readable. It was a nice change from my mysteries and thrillers. In fact, it felt good to exercise my brain and find that learning can once again be fun.

Howard Wainer is an entertaining writer. However, much of this feels like a rehash of things that he has written elsewhere.