07 August 2007

Freakonomics -- Steven D. Levitt and Stephen J. Dubner

The picture is of the revised edition, I read the original edition of this book. After everyone telling me I should read this book and thus making me want to read it less, I was sitting around the house and it happened to be on my parents' bookshelf. So, I decided I would pick it up read it because it didn't look to long and the print was pretty big once I actually opened it. It took less than a day to read and was actually pretty entertaining. However, there are no deep truths in here or really anything too surprising. The central theme of the book seems to be that correlations exist all over in the world, but we must be careful not to confuse correlation and causation. To a scientist/mathematician/economist (I suppose that is what I am), this is an obvious point. But, people who have read this book still make that mistake because they just remember something about abortion, sumo wrestlers, teachers cheating, etc. They do not remember the fine details and hence commit the sin the book seems to be trying to prevent and start to say that X causes Y because they happen to be loosely correlated. I enjoyed the book as something quick and light. I would in fact recommended this book to those who did not study anything difficult in school, i.e. business majors, psychology majors, etc. To someone who knows a lot about the theory of regressions this book is just a nice little compiliation of interesting tidbits, but useless because most people I run into have either read the book or have been told all the little tidbits already.

No comments: