Maarten buis <[email protected]> gives simulation code for calculating
coverage probabilities for "exact" binomial CIs:
> set more off
> capture program drop sim
> program define sim, rclass
> drop _all
> set obs 1000
> gen x = uniform()<.99
> ci x, binomial
> return scalar correct = r(lb)<.99 & r(ub)>.99
> end
> simulate correct=r(correct), reps(10000): sim
> sum correct
A while ago, Nick Cox and wrote -bincoverage-, which will calculate binomial
CI coverage probabilities exactly (and here we do mean "exactly") through the
summing of binomial probabilities.
. findit bincoverage
-bincoverage- allows you to set the sample size, true success probability,
nominal level, and the flavor of exact binomial CI to examine:
Clopper-Pearson (the default), Wilson, Agresti, or Jeffreys.
. bincoverage, n(50) p(0.75) wilson
For N = 50 and p = .75, the true coverage probability of the nominal
95% wilson CI is 0.9519.
--Bobby
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
