Re: comparing two proportions

On Mon, 21 Jul 2008 07:55:18 -0700 (PDT), hvanderklooster@xxxxxxxxx
wrote:

Hello,

I am trying to compare two proportions and I am wondering what
statistical test is best? Here is an overview of my data:

I have:
the number of events pre-intervention
the number of events post-intervention
**the subjects in the pre and post intervention groups are NOT the
same people... the intervention was community-based and we measured
the number of times "x" occurred in the pre intervention time period
and the number of times "x" occurred in the post intervention time
period.

We want to compare the number of events pre intervention to the number
that occurred post intervention such that we can say the intervention
worked. Is comparing the two proportions the best way to do this? Is
there a better way? Also, what SAS procedure would I use?

Sorry to mention it, but you say you have two NUMBERS
without ever mentioning the denominators that would
make these into proportions. If the situation is simple
enough, including equal time periods, then you could assume
Poisson distributions, and compare the two numbers based
on that assumption.

Is this some large sample, with the same subjects and N each time, and
with the same time frame? - You might have a much tougher
argument to make about any "causation" if you are merely
comparing rates in two disparate populations/samples.

Are the events distributed 1-per-person, or is there
dependency? - Using "persons with events" can simplify
this dependency, without much loss if there are very few
instances of two+ events.


--
Rich Ulrich
.

Relevant Pages

  • Re: comparing two proportions
    ... I am trying to compare two proportions and I am wondering what ... **the subjects in the pre and post intervention groups are NOT the ... the number of times "x" occurred in the pre intervention time period ... We want to compare the number of events pre intervention to the number ...
    (sci.stat.math)
  • Re: stats help: comparing two proportions
    ... I am trying to compare two proportions and I am wondering what ... **the subjects in the pre and post intervention groups are NOT the ... We want to compare the number of events pre intervention to the number ...
    (sci.stat.consult)
  • comparing two proportions
    ... I am trying to compare two proportions and I am wondering what ... **the subjects in the pre and post intervention groups are NOT the ... the number of times "x" occurred in the pre intervention time period ... We want to compare the number of events pre intervention to the number ...
    (sci.stat.math)
  • stats help: comparing two proportions
    ... I am trying to compare two proportions and I am wondering what ... **the subjects in the pre and post intervention groups are NOT the ... the number of times "x" occurred in the pre intervention time period ... We want to compare the number of events pre intervention to the number ...
    (sci.stat.consult)
  • =?iso-8859-1?q?Re:_Bob_Ling=B4s_ignorance_is_clear?=
    ... can do in order to compare two INDEPENDENT proportions." ... Fisher's Exact Test is most appropriate for comparing two independent ... be used to compare two independent proportions. ... Not once have I suggested using the Normal approximation to a binomial. ...
    (sci.stat.math)