Customer Reviews:
Showing reviews 6-10 of 110
 Insightful and inspiring September 13, 2009
Allan D. Tokuda
His explanations of "small multiples", the "data-ink ratio", and the proper aspect ratio for graphs have made my academic/professional presentations more effective.
 Towards Legibility Standards for the Display of Data July 1, 2009
Daniel Lobo (Washington, DC More often than not.)
Tufte's volumes is an iconic volume for graphic design that unites legibility with efficiency and beauty in a cogent and stimulating manner. While it's reputation and overwhelming authority may hurt it a bit, it is actually a worth read with a well develop and constructed analysis - with a breakdown of some "best and worst" practices" - and an intriguing attempt to develop a theory for data graphics, which for the most part is worth carrying around and promoting.
Among some of the most obvious defects is a repetition of some trends and examples, a patronizing sense that permeated the narrative here and there, and with that, touches that may seem that the critique is trying to be too final, and offers little room for debate. But all in all these touches are minor and the volume withstands as a pivotal critical point to develop legible and useful graphic information. Something that I would recommend to anyone involved with, or interested in these aspects.
 Dissapointed Scientist June 12, 2009
Science Professor (Oregon, United States)
5 out of 7 found this review helpful
Because of the reviews, I was really excited to read this book. I am a scientist with a graphic arts background that I am told I use to great effect. I was hoping that this book would provide new ideas for presenting my own work and would help me learn to effectively communicate with my graduate students about their graphics. Unfortunately, this book did neither and instead, left me frustrated with the outdated information, overly dramatic characterizations of scientific graphics, and overly reductionist approach.
I could find few graphs that were published after 1980 and not many more than that published after 1960. There are repeated references to drawing plots with a ruler and pen which few, if any, scientists use to produce graphs today. As a result, the information is not of much use. Even the examples used to illustrate the problems with graphics produced electronically look to have been produced with a first generation Mac and bear no resemblence to the scientific graphics of today. While I agree with some of the points made about the graphics used in the text, they are so out of date that it makes it difficult to relate these points to my students.
Another reviewer commented quite clearly about the derogatory tone - with statistics and graphics called "lies" in more places than I can count so I'll leave that point alone.
The biggest flaw for me was the overly reductionist approach used throughout. For example, Tufte presents several alternatives to the venerable box plot. While I agree in principle that reducing useless ink is a good idea, many of these proposed graphs miss the point - conveying information clearly and honestly to the reader. The traditional box plot does a few things very well - it presents the mean of the data as a strong line with the majority of the data presented in an open box, providing visual weight somewhat proportional to the number of data points represented while showing the full range (or some statistical description of the data range) as whiskers which are visually reduced in weight relative to the box, again, relatively proportional to the number of data points they represent. Tufte's proposed alternatives may convey the same amount of data but they are difficult to interpret and do not provide the same visual weight to data but rather provide equal weight to all values represented in the best cases or in the worst, make the viewer focus on the ends of the data distribution instead of the mean value and majority of the data surrounding it. I think this reductionist approach misses how the human brain perceives information in its quest for simplicity and reduced ink. By the time one finishes this book, you'll believe that Tufte pays for graphic ink by the drop. However, text ink is apparently free as the text is often redundant and ovely verbose.
I had hoped to find a book that would be a great resource for myself and my students. I'm still looking.
Don't buy this book. Buy a book on Visual Statistics. May 11, 2009
D. Park
4 out of 8 found this review helpful
I wasted my money on this obsolete book. I truly don't know where these positive reviews are coming from, when the content of the book solely focuses different types of charts, when my 1998 Excel can do a better job plotting a pie chart!
Buy this : Visual Statistics: Seeing Data with Dynamic Interactive Graphics (Wiley Series in Probability and Statistics) This has the most relevant information for someone who is looking for ways to present data accurately and precisely.
Illustrated History of Good & Bad Data Graphics, with Guidelines for Great Graphics. May 5, 2009
mirasreviews (McLean, VA USA)
1 out of 1 found this review helpful
Statistician Edward R. Tufte makes a case for data graphics as respectable tools for representing and understanding data, not dumbed-down pictures for unsophisticated audiences in "The Visual Display of Quantitative Data". Tufte lays out examples of good and bad graphics and presents a "practical theory of data graphics" which the author believes will produce clearer, more informative, and more pleasing graphics than are often seen in publication. Tufte advocates "the simultaneous presentation of words, numbers, and graphics", not just as way of presenting information, but as facilitating other ways of understanding data. This second edition of the book adds color to some of the hundreds of graphics that Tufte uses to illustrate his ideals.
The subject is presented in two parts: The first addresses Graphical Practice and the second Theory of Data Graphics. Graphical Practice begins with a chapter that presents and discusses examples of excellent data maps, time-series, and spatial time-series over the past few centuries. Many examples from the work of 18th century political scientist William Playfair are featured, as they were pioneering and quite beautiful. Next is a chapter presenting examples of bad graphics, graphics that deceive, and conclusions on how to avoid these pitfalls. That is followed by discussion of the reasons we have so many misleading and dumbed-down graphics.
The Theory of Data Graphics addresses different aspects of design, as the author prescribes the maximization of the "data-ink ratio", meaning that a high proportion of your ink should be dedicated to the data itself, and proscribes "chartjunk", or extraneous decoration, such as moiré effects, that tend to consume and confuse the data. Tufte also talks about multi-functioning graphical elements, data density, and proportion. My own view is that replacing the frame of a scatterplot with a range-frame, which Tufte advocates, leaves the viewer without a point of reference and inhibits comprehension. Like all ideologies, Tufte's can be taken too far. But "The Visual Display of Quantitative Information" will help graphic artists, students, and professionals make the most of their charts, graphs, or maps. And it showcases a lot of creative and lovely graphics.
Showing reviews 6-10 of 110
|
