Mag,

The theory of statistics states that random sampling is representative
of the population as a whole, so no, you don't need a large number of
people to exterpolate findings.  A telephone survey, for example is
random sampling as long as you randomly pick numbers (and
computer-generated sampling ensures this) and so the results of a poll
can be said to be representative of the population.

This of course, depends on how the questions are stated.  They have to
be neutral, not slanted.

I know all of this because 1) I took several statisitcal  analysis
courses in college and grad school, and 2) I worked for a marketing
research firm back in the 1980s as a supervisor who ensured quality in
coding the raw data.

The rallies are not truly representative of the population at large
because they only attract people who are against the war and want to
express it in that fashion.  There are likely many more against the war
who don't want to march, as well as many in N. Calif. who are for the
war but don't go to marches.  Peace marching is *not* a truly
representative indicator of the US (or even Calif) population as a whole.

Remember that Californians are the one who elected Reagan to
governorship and the ones who sent him to the Presidency.  'Nuff said.
I'd venture to say Californians are much more conservative as a whole,
than outsiders think.


Kat

Magenta Raine wrote:

>Which poll was this?  Also, only 1000 people were polled. That's hardly
>enough people. We had 8,000 here in Oakland on Saturday.
>
>Mag
>
>
>