Re: [TCLUG:19672] VALinux vs Compaq, and OT: statistical inference

[Mike Hicks <hick0088@tc.umn.edu>]

Philip C Mendelsohn wrote:
> 
> On Tue, 18 Jul 2000 andy@theasis.com wrote:
> 
> > > statistics *never* imply causality, nor guarantee outcomes.
> >              ^^^^^^^^^^^^^^^^^^^^^^^
> >              That's not true              ^^^^^^^^^^^^^^^^^^
> >                                          that is.
> 
> I'm not going to argue the point.  Look in any statistics book at your
> convenience.  What I said is provably correct, so I have no bones to pick
> about it.~

I think the key word is `imply'.  Unless your definition is different
than mine, you can imply a lot of things, but you can't prove nearly as
much.  Of course, I'm weird, since I think `proof by induction' is
roughly equivalent to `proof by implication'...

-- 
 _  _  _  _ _  ___    _ _  _  ___ _ _  __   Air conditioned 
/ \/ \(_)| ' // ._\  / - \(_)/ ./| ' /(__   environment - Do not  
\_||_/|_||_|_\\___/  \_-_/|_|\__\|_|_\ __)  open Windows. 
[ Mike Hicks | http://umn.edu/~hick0088/ | mailto:hick0088@tc.umn.edu ]