On Sat, 11 Nov 2006 16:12:34 -0800, "Peter Duniho"
wrote:

"Jessica Taylor" wrote in message
...
Not true. The very foundation of Statistics is infering facts about an
entire
population through the use of a much smaller representative sample.

The "facts" that one infers describe a known *portion* of the population.
You cannot prove anything about 100% of a population with statistics. It is
impossible.

Not true. We can prove that a medicine is effective at treating a malady
without testing its effectiveness on 100% of its population.

Isn't that "We can show within a specific margin of error
(probability) what we can expect from the general population?"


You cannot use statistics to prove that a medicine will be effective on 100%
of the population.

The two of you are using different definitions of proof.
Stastical proof and absolute proof.

When it comes to stastics there are usually exceptions to a proof.
It's sorta like side effects from medication. Even when less than the
placeibo they still have to be listed which generally means those side
effects were probably "all in their head" or elsewhere.



We can prove that
chronic smokers have a higher risk of bladder and lung cancers than their
peers
who have never smoked without needing to find the entire populations of
smokers
and non smokers.

The very term "risk" precludes an absolute statement about 100% of the
population. That's why statistics can be used to describe risk. You can
use statistics to prove "risk", but you cannot use statistics to prove
actual outcomes. In this example, you cannot use statistics to prove that
100% of chronic smokers WILL have bladder and lung cancers. The best you
can do is prove that it is likely a certain percentage of them will.

Those are two different things. If you fail to comprehend that, you have no
idea what statistics actually is.

Statistics has nothing to do with it.

Not true but then again, I don't believe that the original poster said
100%.

Any blanket generalization is necessarily applied to the entire population.
That's why blanket generalizations are so offensive in the first place.

I'm not surprised you're having trouble understanding this. You're a woman.
Those people are the least mathematically inclined around!

Pete

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com