On 6/25/2018 11:40 AM, Dan Asimov wrote:

----- Whenever you see some statistical correlation used to imply causation ... -----

Correlation alone should *never* be used to infer causation.

It's not only difficult to infer causation, it's difficult to define.

But you could  have determined the age to subjects, had them listen to "When I"m Sixty Four", and determine that they were three minutes older.  Wouldn't you then be able to infer (I used the weaker "imply") that listening to "When I"m Sixty Four" causes you to be older? Brent

—Dan

Brent Meeker wrote: ----- Of course the study that shows that listening to "When I"m Sixty Four" makes you younger is one that shouldn't require statistical correlation analysis at all. Whenever you see some statistical correlation used to imply causation the first thing you ask yourself is, "Could this have been tested directly?" There's no point trying to clean up statistical inference if none is needed in the first place. -----

_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun