Re: correlation, critical temperature

[John Bois <jbois@umd5.umd.edu>]

On Sat, 20 Jun 1998, Paul Willis wrote:

> 
> >>I reiterate that every correlation implies a causal link...
> 
> This is one of the great fallacies of logical thinking.
> 
> Correlation DOES NOT imply causal link

And even seemingly obvious causal connections can be totally wrong.  For
example, as a teacher in U.S. high schools I argue that out-of-control
students are, well, out-of-control.  The only available remedies to this
are suspension (totally counter-productive) or ritalin prescription (a
cure for childhood!).  I argue for corporal punishment because it rarely
has to be used.  But, back to the point, people are always citing studies
"proving" it leads to violence later in life.  Such studies ignore the
likely underlying cause, viz, pain-in-the-ass adults were probably
pain-in-the-ass kids.  In other words these studies cannot measure
underlying (and probably more robust) genetic causes.   And yet, social
"scientists" are completely oblivious to this possibility.