[extropy-chat] The NeoCon Mind-Trick (was letter concerningpresidential growth)

[The Avantguardian]

--- Joseph Bloch <transhumanist at goldenfuture.net>
wrote:

> Unfortunately, this is also not valid, for the
> simple reason that you 
> are trying to compare two completely different
> populations.

Actually we are not trying to compare two different
populations so much as trying to determine the degree
of dependency between two populations. That is to say
we are looking to see if there is any overlap between
two presumably separate populations. This is every bit
as valid as taking a population of smokers and a
population of people with cancer and trying to see if
there is any correlation between the two. Obviously
not everybody who smokes gets cancer and not everybody
who gets cancer smokes, but if there is any overlap
between the two then you should be able to find it.
 
> That is, persons who voted in the 2004 Presidential
> election are not the 
> same as college-bound 18-year-olds from 1998.

The two population are not identical but it is highly
likely that some proportion of the people voted in the
2004 Presidential election in the same state they took
their SATs in 1998.  

> Maybe
> this is only clear 
> to me because I work with election polls for a
> living, but that's a 
> major no-no.

Polls are different statistical beast. In a poll you
take a representative sample of a population and try
to extrapolate data from the sample to a population.
Here what Spike and I are doing are taking two
different populations that is to say ALL the people
that took the SAT in a given year in a particular
state and ALL the people that voted in a given year in
particular state and seeing if there is any overlap
between the two populations. The statistics of this
are more akin to epidemiology than they are to
surveys. 

 In order for this to be really
> statistically valid, you 
> would need to do some sort of exit polling from
> representative samples 
> throughout the country, get some sort of
> standardized IQ data for those 
> people, and make the correlation from that. Not
> gonna happen.

Why deal with the uncertainty introduced with
representaive statistical samples when you can deal
with entire populations?  

The Avantguardian 
is 
Stuart LaForge
alt email: stuart"AT"ucla.edu

"The most beautiful thing we can experience is the mysterious. It is the source of all true art and science. He to whom this emotion is a stranger, who can no longer wonder and stand rapt in awe, is as good as dead: his eyes are closed. . ."

- Albert Einstein, "What I Believe" (1930)

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com