Dear Allstat:
Is there a statistical test that can be used to compare two proportions, p1
and p2, where the sample from which p1 comes is a fairly small subset of the
sample from which p2 comes? In one case, the sample corresponding to p1 has
very different characteristics than the much larger sample (p2's sample) to
which it belongs (things are 'lumpy'). In the other case, the sample
corresponding to p1 may or may not be all that dissimilar from the larger
sample to which it belongs. Note that p1 and p2 are not the same proportion,
i.e. the two are capturing different phenomena
This is not a situation in which the two samples are paired in any way.
Hope this makes sense. I have been unable to find anything out there
except the standard z-test for independent samples, McNemar's test for
paired samples. The two samples involved here are neither independent nor
paired.
Thanks very much for any help you can give,
Marla
[log in to unmask]
|