The purpose of the original post was not to argue over WHY home-schooled
kids do markedly better on standardized academic tests, but rather to refute
with this statistical fact the ignorant attempt at slandering home-schooling
as some sort of fringe cult whose students are profoundly inferior
performers. The statistics DO clearly show - for whatever reason - that this
ignorant prejudice is clearly false.


"Ed Huntress" wrote in message
et...
"Bray Haven" wrote in message
...

Kids from disfunctional families and illiterate, impoverished families
are
included in the "general population." Home-schooled kids almost
exclusively
are not.

Another case of lying with statistics.

--
Ed Huntress

That may be true in general but I live in a very low income county and
many low
income families home school very successfully. I don't know that I'd
lump
them
with illiterate or even dysfunctional. "Lying" is probably a bit of an
exageration but then exagerating is lying too, isn't it )
Greg Sefton

I don't doubt that home-schooled kids do quite well on tests, Greg. It
appears that the average home-schooled kid actually is advantaged, though,
in several ways. They have parents who really care; they have parents who
make substantial sacrifices to educate their kids; they get something
close
to one-on-one interaction with the teacher, full-time. What's the per-hour
value of a parent capable of teaching, 3 or 4 hours/day? How does that
stack
up against the cost of public education? My guess is that it works out to
around $15,000 kid for home-schooling, versus something less than
two-thirds
that for public school. More money, better education, eh?

The advantages are great in themselves, and I have nothing against
home-schooling. But it makes the statistical comparisons flaky or
misleading. And a lot of home-schooling advocates aren't above using the
statistics to draw a conclusion that's opposite of the truth. It's hard to
believe that all of the people who quote these stats are so cavalier about
them, or so ignorant of statistical methods that *all* of them fail to see
the errors.

Ed Huntress