David:
I'll jump in before the wise people do. It comes from Sir Francis Galton,
correlating (say) height of parents and children. Tall parents have taller than
average children, but the children are not as tall as the
parents. Short parents have shorter than average children, but the children are not as
short as the parents. Both sets of children "regress" toward the population mean.
Tryon found the same thing with maze bright and dull
rats. The offspring were never as bright (or dull) as their parents. Gonick and
Smith's "Cartoon Guide to Statistics" says Galton also used the phrase "regression
toward mediocrity" - and this may explain why he chose a
word with negative connotations. Perhaps he saw the ends of the distributions (esp.
with IQ) as an advancement. Movement back to the population mean would be "regression"
toward mediocrity.
I believe the negative connotation is inappropriate. By definition, as the
correlation between 2 variables decreases, your predictions must regress into the mean
of the Y variable (Y predict = Y bar when r =0). It
doesn't mean the height of children becomes less variable or that successive
generations were moving toward homogeneity. (Galton was smart enough to know this).
On a related note, there is something called a "regression fallacy" (Freedman,
Pisani, Purves, etc _Statistics_). If you test the same people twice - those with the
highest scores will tend to have lower scores on the
retest, and those with the lowest scores will tend to be higher on the retest. This is
inevitable. But it's a fallacy to believe there is a force at work reducing
variability, pushing scores inward, and making people more
similiar to each other.
DAVID KREINER wrote:
> One of my students asked why the statistical process of regression is called
>regression. I could not come up with a plausible answer, so I'm asking my wise
>Tipster colleagues to help. Does anyone have a good answer?
--
---------------------------------------------------------------
John W. Kulig [EMAIL PROTECTED]
Department of Psychology http://oz.plymouth.edu/~kulig
Plymouth State College tel: (603) 535-2468
Plymouth NH USA 03264 fax: (603) 535-2412
---------------------------------------------------------------
"What a man often sees he does not wonder at, although he knows
not why it happens; if something occurs which he has not seen before,
he thinks it is a marvel" - Cicero.
