"The Commonsense Guide to Understanding Numbers in the News, in Politics, and in Life" is a new book with a longwinded title that attempts to do exactly what it says. The authors (Michael Blastland and Andrew Dilnot) take all those numbers we see on the news -- those misused statistics that are so attractive -- and debunk them.
For instance, consider a news story that says there is a correlation between big hands and better reading comprehension. This is a classic case of two factors being correlated, but not causally related. It may be that these two things go together, but having big hands does not cause better reading. And so they found. Bigger people have bigger hands. Adults are bigger people. Adults are better readers than children, who have small hands.
Or the study that showed Republicans enjoy sex more than Democrats. Good dating tip, I guess. But more men vote Republican than Democrat. And men claim to enjoy sex more than women.
I haven't read the book yet. These examples are from reviews I've seen. The book sounds like the flip side of Steven Levitt's "Freakonomics," which I read a couple of years ago. In Freakonomics, Leavitt explains things like why legalization of abortion has reduced the crime rate.
So in Freakonomics, sometimes things that don't seem to be related, actually are. In the Commonsense Guide, things that seem related, actually are not.