Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: RE: a "byable" cii

From   "Nick Cox" <>
To   <>
Subject   st: RE: RE: a "byable" cii
Date   Thu, 6 Feb 2003 20:40:44 -0000

Sayer, Bryan
> Gee, personally I would suggest just doing them in the 
> spreadsheet, if that
> is the way the data came.  (Standard error of proportion = 
> sqrt[p*(1-p) / n]
> , 95% c.i. = p +/- 1.96*se(p))

That is a common textbook formula, but it 
behaves much worse than is widely realised, 
even if p is not close to 0 or 1 and n 
is moderately large. 

Much recent literature (some 2001 or 2002) shows how lousy 
it is, and how there are much better alternatives, 
which do _not_ include the so-called "exact" 
Clopper-Pearson interval implemented by -ci- 
and -cii- for binomial proportions. ("Exact" is 
a propaganda term here.) 

Some alternatives are implemented by -cij-/-ciji- 
and -ciw-/-ciwi- on SSC. The help files for 
those packages -- which have just been updated -- 
contain much more detail and several references. 

The approach suggested in my reply to Paul 
can readily be extended to use either -ciji- 
or -ciwi-. 

(The latter has nothing to do with New Zealand.) 


*   For searches and help try:

© Copyright 1996–2019 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index