well, let's see
if you take ANY sized sample from a target population ... as long as you
don't have all the elements in your sample (ie, you don't actually have the
population) ... then conceptually there is a standard error
that is, for that sample size ... if you repeatedly take samples and
calculate the statistic of interest (say a proportion of something in your
samples) ... and then make a distribution of all those sample statistics
(ie, those sample proportions) and calculate the standard deviation of that
SAMPLING DISTRIBUTION of proportions ... THAT is your standard error
of course, as a practical matter, there are at least two considerations
related to sample size
1. assuming that your population is relatively large, then the larger your
sample, the closer will be your statistic TO the parameter of interest and
therefore, the smaller will be your standard error
2. as the proportion gets larger of the size of your sample is to the size
(if known) of the target population, then the more your sample data
approximate the population data and, the smaller will be your standard error
in most sampling cases, #2 is not of too much importance since, we usually
are only able to sample a small fraction of the target population size but,
IF the target population (by definition) is small, then one wonders why
there is really much of a sampling issue involved since, it seems like that
the total population (ie, census) could be obtained ... or close to it
At 12:39 PM 11/30/00 +, [EMAIL PROTECTED] wrote:
>Hello,
>
>who knows an answer to the following question?
>Let's assume that I take a sample of e.g. 100 people. I ask them a
>question and, e.g. 50% say "yes". I construct a 90%-confidence interval
>and get a standard error of 11.6. Fine.
>
>However, this assumes that the population size is unlimited.
>If the population was 100, then I would not have any standard error.
>However, what if the population size is 200? How do I construct a
>confidence interval then?
>
>Anybody with an idea?
>
>Thanks in advance
>
>Nils Riemenschneider
>
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.
>
>
>=
>Instructions for joining and leaving this list and remarks about
>the problem of INAPPROPRIATE MESSAGES are available at
> http://jse.stat.ncsu.edu/
>=
dennis roberts, educational psychology
penn state university, 208 cedar building
university park, pa USA 16802 ... AC 8148632401
[EMAIL PROTECTED] ... http://roberts.ed.psu.edu/users/droberts/drober~1.htm
=
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=
