How to lie with and without statistics

[quark]

Does anybody else think that "How to lie with statistics" should be the book on the bottom? If somebody wants to lie with statistics, he/she would have to learn statistics first, that's more, not less.

[Jake]

I'm not sure if the picture makes sense. I guess there are two ways to interpret the size of the books: (a) bigger book about X = more ways to do X, (b) bigger book about X = more difficult to do X.

I guess (a) can't be right, because clearly there cannot be fewer ways to lie in general than there are ways to lie with statistics or lie without statistics (the latter two are subsets of the first).

But (b) doesn't seem right to me either. It would mean first that it is harder to lie using statistics than it is just to lie in general -- fair enough I think (see also my signature!!) -- but then it is also harder to lie without statistics than it is to lie with statistics? That doesn't seem consistent.

Where is this picture from? What is it supposed to mean exactly?

[noetsi]

I think the point is that it takes more effort to lie with statistics. Lying is simpler and thus requires less effort and details.

I dont think you can actually lie with statistics. Because it can not establish causality under any circumstances which rules out lying. What you can do is confuse the heck out of people and yourself. Another reason to make the statistics book longer.