[R] binom.test

[Ethan Johnsons]

Thank you for the info.  It helps.

After all, it would be:

> 0.1304348-1.96*(sqrt((0.1304348*(1-0.1304348))/46))
[1] 0.03310968
> 0.1304348+1.96*(sqrt((0.1304348*(1-0.1304348))/46))
[1] 0.2277599

Does R have a function for the calculation above?

ej


On 10/20/06, Francisco J. Zagmutt <gerifalte28 at hotmail.com> wrote:
> Ethan,
>
> You need to explain why you think this is "not the right function to use". R
> is doing exactly what you are asking it to do.  Now is up to you to choose
> the methodology you feel is correct.
> For a good discussion on your particular issue I recommend you the following
> reference:
>
> A. Agresti and B. A. Coull, "Approximate is better than "exact" for interval
> estimation of binomial proportions," The American Statistician, vol. 52, no.
> 2, pp. 119–126, 1998.
>
> Once you figure out the "right function to use" see if the function is
> available in R.   If not readily available, and if after searching through
> R's documentation and the forum archives you still can't find a way to
> perform the calculation, then is time to get back to this forum.
>
> Regards,
>
> Francisco
>
>
> Dr. Francisco J. Zagmutt
> College of Veterinary Medicine and Biomedical Sciences
> Colorado State University
>
>
>
>
> >From: "Ethan Johnsons" <ethan.johnsons at gmail.com>
> >To: r-help at stat.math.ethz.ch
> >Subject: [R] binom.test
> >Date: Fri, 20 Oct 2006 17:18:02 -0400
> >
> >A quick question, please.
> >
> >46 e coli lab samples are tested,  6 of them returned positive.
> >
> >So, the best point estimate for p is  6/46 = 0.1304348.
> >
> >For a 95% CI for p,  I thought binom.test would give me the correct
> >result, but it seems it is not the right function to use.   What is
> >the R function for this?
> >