Xiao Li <[EMAIL PROTECTED]> wrote in message


news:[EMAIL PROTECTED]


Let N be the population size.  Let p denote the proportion of this
population with a certain characteristic (ie. has diabetes, divorced,
homosexual, etc.)  Let's say we are trying to estimate p using a
sample of size n.  So we collect our sample and calculate p^ to be the
proportion of elements in our sample with our characteristic of
interest.  Then we calculate the standard of deviation of p^ (also
known as the standard error) because it gives us an idea of how close
our sample proportion is to the real proportion.

Now, my statistics textbook says the standard of deviation is given by
SD(p^) = root[(p^(1-p^))/n] when sampling with replacement or when
the sample size is significantly smaller than the population size.

In other words, when n << N, the effect of N is negligible on the total SD. So N _is_ included in the equation, but it has little effect. If you work out the hypergeometric form of the equation, and then run N up very large, for constant n, you will see that the SD approaches its value (determined without N, above) asymptotically . Try it with what Joe said.

Jay


My question is why isn't there a N variable in our equation for
SD(p^)? According to this formula, if we are trying to figure out the
proportion of people with diabetes in Berkeley, California with sample
size n, we would get the same standard of error as if we are trying to
figure out the proportion of people with diabetes in the world with
sample size n. (In the former, the N is a lot smaller than in the
later).



I think this is a case of the hypergeometric distribution, instead of
binomial. The hypergeometric takes into account cases of
non-infinite N.
------ Joe ------- [EMAIL PROTECTED]


--
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