[R] Exact conf. interval for one proportion

Heberto Ghezzo heberto.ghezzo at mcgill.ca
Tue Sep 11 14:44:48 CEST 2001


Just in the last American Statistician a paper by Robert G. Newcombe(AS 55-3 Aug
2001 p200-2
"The Wilson limits for logit(pi) can be expressed as
logit(p)+- arcsinh(z/(2*sqrt(n.p.q)) He refers to his paper in Stats in medicine
1998-17-857-872 in which he finds by simulations that the coverage of Wilsos's in
952 while the Clopper is .971 and Walds .881
Wilson's by the way is from 1927 so newer is not always better. . . .

> >>>>> Luis A Izquierdo writes:
> > Dear R people:
> > I need to estimates some confidence intervals for one
> > proportion. Reference for binom.test in ctest package, I suspect is so
> > old and perhaps now there are better methods. Wrong or not?  Please, I
> > need your advise.

r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list