You have to check also for the total number of subjects in the sample. If
there are less than 40 subjects, the estimator suggested by Brown, L. D.,
T. T. Cai, and A. DasGupta. 2001. (Interval estimation for a binomial
proportion. Statistical Science 16: 101-133.) is the Jeffreys one, instead
than the Agresti.
best
marco braggion
On Sun, Mar 11, 2012 at 4:52 PM, Michael Baxter <
[log in to unmask]> wrote:
> You apply a suitable transform to your data, calculate the confidence
> interval on the transformed data and then invert the transform. Michael
> Baxter >>>> Dan Abner <[log in to unmask]> 3/7/2012 1:10 PM >>>
> > Hi everyone,
> >
> > I am attempting to construct 95% ci's for the regional means of a set
> > of proportions. So there are anywhere from n = 1 to 12 proportions per
> > region and I want a ci for each regional mean. However, the upper
> > bounds of the ci's sometimes exceed 1.000.
> >
> > Can anyone suggest a good method for bounded ci's for this application?
> >
> > Thanks!
> >
> > Dan
>
> You may leave the list at any time by sending the command
>
> SIGNOFF allstat
>
> to [log in to unmask], leaving the subject line blank.
>
--
http://taffey6977.blogspot.com/
http://www.sentireascoltare.com/marco.braggion
http://www.mangialibri.com/taxonomy/term/175
http://www.facebook.com/taffey6977
http://www.anobii.com/taffey6977/books
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|