I read the second Tufte book I bought, Visual Explanations. I previously wrote my thoughts on The Visual Display of Quantitative Information.
It's second chapter, Visual and Statistical Thinking: Displays of Evidence for Making Decisions, is probably the best. It illustrates how having good displays of evidence can be vital for making good decisions and thinking about problems, with two particular case studies:
- the cholera epidemic in London in 1854, where John Snow performed and outstanding job in finding the source of the disease, and
- the failed launch of the space shuttle Challenger in 1986, which exploded shortly after launch because the temperatures in the day it was launched were outside of the normal operating parameters of some of its parts.
While the analysis of both cases is interesting and well researched, it didn't really convince me. It's overly simplistic to conclude that Snow succeeded and the NASA officials failed simply because of the use or lack of use of good graphs, rather than because of solic scientific thinking. I think the author is somewhat aware of that, but that doesn't stop him from implicitly attributing success and failure to the use of “good” graphs.
I also was annoyed to see the author criticising Feynman's presentation —it describes it as “deeply flawed”— totally missing its point, but I suppose this is not very important.
Then we find a chapter coauthored with a magician, going over diagrams of books describing magical tricks. While at first this may seem an interesting idea, I found it also relatively unconvincing. I see little point to this, which I found pompous rather than rigorous.
There are other chapters, all unremarkable. The last one, Visual Confections, is specially bad, just a potpourri of different drawings from different centuries —from the beginning of the XVII century to some computer screens— with some almost-arbitrary positive or negative criticism with very little point or underlying structure.
Even though the book is well written and the effort that the author put in making the form of the book friendly (such as repeating a drawing so you won't have to go back to the previous page to see it again) does show, I struggled to finish it. In the end, I think it:
- lacks an underlying structure or theory, being just a random assortment of images and comments which, while compelling at first, quickly becomes boring, and
- just plain isn't very convincing.
I wouldn't recommend it.
Discussion (1)Last update: 2010-01-22 (Rev 16620)