Other way around.  If the sample size is greater than about 1/10 of the total
N, the confidence interval becomes smaller.  If we sample the entire
populaiton, the CI is +/- 0.  (assuming the measurement was precise, too:)

Jay

Fred Ettish wrote:

> "Jerry Dallal" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
> > Fred Ettish wrote:
> > >
> > > A recent poll by the Pew Group of 15000 Muslims world wide found 71%
> support
> > > Bin Laden. Looking at my textbook the formula for a population
> proportion
> > > is:
> > >
> > > p +/- Z * SQRT ( p(1-p)/n )
> > >
> > > Ok simple enough, but my question is doesn't the population size affect
> the
> > > confidence interval? How does the fact that their are 2000000000 affect
> the
> > > interval?
> > >
> > > Thanks
> >
> > It doesn't.  When the sample size is, say, less than 10% of the
> > population size.  The uncertainty reflected in the CI is the
> > uncertainty of the measuring instrument, that is, the point estimate
> > from the sample.
>
> So if a sample of 15,000 came from a population of 15,001 one couldn't be
> confident about a narrower interval than if it came from a population of
> 10^10?
>
> .
> .
> =================================================================
> Instructions for joining and leaving this list, remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives are available at:
> .                  http://jse.stat.ncsu.edu/                    .
> =================================================================

--
Jay Warner
Principal Scientist
Warner Consulting, Inc.
4444 North Green Bay Road
Racine, WI 53404-1216
USA

Ph: (262) 634-9100
FAX: (262) 681-1133
email: [EMAIL PROTECTED]
web: http://www.a2q.com

The A2Q Method (tm) -- What do you want to improve today?



.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to