Superfreakonomics by Steven D. Leavitt and Stephen J. Dubner

I read Freakonomics (or maybe I listened to it, I can't remember) and liked it, and so I figured eventually I'd read Superfreakonomics and that I'd probably like it in about the same way. And I was right.

I find it hard to say much more about it. There's not really a general theme to the book, other than that data analysis can turn up surprising patterns and conclusions about interesting questions. The book is well-researched, jauntily (but not annoyingly) written, and nicely blends broad issues, number-crunching, and stories of individuals.

I think maybe my favorite chapter, and the one that exemplifies what I like about the book, was the chapter that began with the case of the callous New York City neighbors. It's the story everyone has heard, in sociology classes and pop science podcasts, about the woman who was murdered while dozens of neighbors looked on and did nothing. Levitt and Dubner revisited the story and essentially reconstructed the events, painting the scene in an entirely new (and much more plausible) light. Conventional wisdom always comes from somewhere, and in this day and age we're prone to attribute much of our conventional wisdom to scientific study. But the world is a complicated place, and the study on which we place our faith one day can be thrown into question or seen from new perspectives or reevaluated. Faith in science can also be misplaced.

Of course, Levitt and Dubner's intention was not to discredit science, but rather to turn a more critical eye on the science of human nature and encourage a finer grained analysis. It's the kind of smart reading of data that I try to develop when I teach basic statistics in my courses, a recognition that the world is a complex place and that everything is always worth a second look.