stat question: comparing proportions in 2 dependent samples

I'm looking for an appropriate statistical test for the difference in
proportions. I want to compare the proportion of a factor (for example,
males) in population A versus the proportion in population B.
However, population A is a partial subset of the larger population B (i.e.,
most, but not all, of population A is contained in population B).
I can't use the ChiSquare test as the 2 samples are not independent (such
as cases vs controls). I'm reluctant to use the McNemar's test for
dependent proportions as the 2 samples are not paired (such as before vs
after in the same population). Can anyone recommend a suitable statistical
test for the difference in these proportions? I'd be delighted with any SAS
code as well!
0
wcw2 (31)
7/17/2007 5:58:52 PM
comp.soft-sys.sas 140164 articles. 1 followers. Post Follow

2 Replies
670 Views

Similar Articles

[PageSpeed] 26

Take a look at the proc freq documentation.  One thing you will find is:

McNemar's test
Beginning in Release 6.10, use the AGREE option in Base SAS PROC FREQ.
Before Release 6.10, create a three-way table with a stratum variable
identifying each subject (or matched group), a variable indicating each
occasion (condition or individual within matched group), and a binary
response variable. Then use the CMH option. For example, if each subject
gives a binary response to each of two drugs, use the statement:
  tables subject*drug*response/cmh2 noprint;

HTH,
Art
--------
On Tue, 17 Jul 2007 13:58:52 -0400, wcw2 <wcw2@CDC.GOV> wrote:

>I'm looking for an appropriate statistical test for the difference in
>proportions. I want to compare the proportion of a factor (for example,
>males) in population A versus the proportion in population B.
>However, population A is a partial subset of the larger population B
(i.e.,
>most, but not all, of population A is contained in population B).
>I can't use the ChiSquare test as the 2 samples are not independent (such
>as cases vs controls). I'm reluctant to use the McNemar's test for
>dependent proportions as the 2 samples are not paired (such as before vs
>after in the same population). Can anyone recommend a suitable statistical
>test for the difference in these proportions? I'd be delighted with any
SAS
>code as well!
0
art297 (4213)
7/17/2007 11:10:00 PM
wcw2@CDC.GOV wrote:
>
>I'm looking for an appropriate statistical test for the difference in
>proportions. I want to compare the proportion of a factor (for example,
>males) in population A versus the proportion in population B.
>However, population A is a partial subset of the larger population B (i.e.,
>most, but not all, of population A is contained in population B).
>I can't use the ChiSquare test as the 2 samples are not independent (such
>as cases vs controls). I'm reluctant to use the McNemar's test for
>dependent proportions as the 2 samples are not paired (such as before vs
>after in the same population). Can anyone recommend a suitable statistical
>test for the difference in these proportions? I'd be delighted with any SAS
>code as well!

Why?

I'm not being frivolous.  The underlying reason for this is likely to
determine
whether this is statistically appropriate or not.  Either way, you ought to
be looking at sample B vs. (population A - sample B).

If you have two proportions which come from different samples, but
there is an issue with the populations that the samples come from, then
that is a completely different question.

So I'm not clear on exactly what you want.

HTH,
David
--
David L. Cassell
mathematical statistician
Design Pathways
3115 NW Norwood Pl.
Corvallis OR 97330

_________________________________________________________________
http://imagine-windowslive.com/hotmail/?locale=en-us&ocid=TXT_TAGHM_migration_HM_mini_2G_0507
0
davidlcassell (5567)
7/26/2007 9:16:39 PM
Reply: