Quote:
Originally Posted by jamesqf
Yeah, that was my point: you can get any stupid correlation you want, if you carefully select the data - as for instance choosing exactly who you'll define as a 'pirate' - to produce the results you want.
|
I remember an episode of "Square One" where the team on Mathnet investigated claims that eating chuckberries caused acne and determined that the chuckberry eaters were teenage boys while the control group was old people.
That is what you are saying, right?
