🚀 go-pugleaf

In article <C17z6J.1MHD@austin.ibm.com> me@vikings.austin.ibm.com (Michael E) writes:
>Your "fact" does not address the problem.  

Actually, your refutation is far more interesting than my factoid ...

>First, the statistic applies to the current American schooling situation. 
>The amount of students requesting entry into private schools will drop 
>acceptance rates to unpopular levels under the voucher system. 

Yes, that's true.  And it's true because people will flock, in greater
numbers than before, to private schools.  Fixed quantity of something
or other divided by large quantity of consumer means a smaller fraction
of the consumers get the good or service.

>Second, the statistic does not mention how someone arrived at that figure. 
>The following lists some questions any good debater or statistician would 
>want answers to:
>o	What popuolation of schools did the researcher cover?
>o	Does the figure represent an average? If so, which average (e.g.
>	mean) did they use? What standard deviations did they come up with
>	and why did they not list them?
>o	Do the population of schools inquired live in the same environment
>	(e.g. same state laws, funding)?
>o	Where does the data come from? If from other sources (e.g. the schools
>	themselves) then how reliable are they? How consistent do the 
>	external sources collect their data? 
>o	Define applicants. Define accept.
>o	Where did you get the statistic?

At the 98% level there isn't much room for deviation.  Let's say that
the standard deviation is 10%.  Oh wait, it can't be 10% because you
can't have more than 100% acceptance and that would mean the data is
so skewed towards 98% that the standard deviation would be about 1%
anyhow.  Hmmm.

Next is the school populations.  See rules about selecting valid
statistical samples and the applicability of small samples to larger
populations.

Next is the sample area.  Actually, it was the worst possible population.
Poor minorities.  Presumably rich non-minorities get in at least as
often as poor minorities.

>Don't use stats unless you give the entire picture.

Get a clue.
-- 
John F. Haugh II                  [ PGP 2.1 ] !'s: ...!cs.utexas.edu!rpp386!jfh
Ma Bell: (512) 251-2151           [ DoF #17 ]        @'s: jfh@rpp386.cactus.org

The DDT ban has killed more Africans than the AIDS epidemic.  Legalize DDT now.