
Concerns • 473
a change in magnitude to another change in magnitude. We nd Mill’s
second statement to present a potential argument for causation. As we
can see, proving causation is a nontrivial task. A reasonable discussion of
Mill’s canons is available at http://www.philosophypages.com/lg/e14.htm.
Ludic Fallacy
When the ludic fallacy occurs we might assume we have a perfect statis-
tical model, or one without a aw, when, in fact, we do not (we are not
sure such a beast can even exist!). Nassim Taleb coined the term in e
Black Swan.
9
In short, we may see a plethora of real situations where the
independently distributed data concept is violated. ...