Nice example. Perhaps I should have said "partial (and not necessarily
linear) correlation."
----- Original Message -----
From: "jim clark" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, December 04, 2001 1:37 PM
Subject: Re: Who said "Correlation does not imply causation".
> Hi
>
> On 3 Dec 2001, Karl L. Wuensch wrote:
> > I think that phrase has created much misunderstanding. I try
> > to convince my students that correlation is necessary but not
> > sufficient for establishing a causal relationship.
>
> And I teach that NEITHER presence NOR absence of _simple_
> correlation can be used to infer much about causality. The
> reasoning is the same in both cases. In the case of presence of
> a correlation between X and Y, another variable might be
> confounded with X and responsible for the observed r. In the
> case of the absence of a correlation between X and Y, another
> variable might be confounded with X and masking its influence,
> producing the observed r.
>
> An example of the latter that I am fond of is the relatively low
> r between amount of time spent studying and grades. The effect
> of study time is masked by its negative correlation with ability
> (i.e., IQ), which tends to be positively correlated with grades.
> In essence, brighter people spend less time studying (perhaps
> because they can obtain good grades without more study time).
> Multiple regression with both study time and ability reveals the
> robust effects of study time (and enhances the effect of ability
> as well).
>
> So, depending on precisely what Karl means by "correlation is
> necessary," I'd have to disagree strongly.
>
> Best wishes
> Jim
>
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
