abd (version 0.1-21)

wilsonCI: Wilson (Agresti-Coull) CI for a Binomial Proportion

Description

Calculate the confidence interval for a "small sample" binomial proportion using the Agresti-Coull (1998) method. Useful for small samples where the Wald "large sample" normal test used by binom.test is not appropriate (when n is small or when p is close to 0 or 1).

Usage

wilsonCI(x, n, conf.level = 0.95)

Arguments

x

an integer number of successes

n

an integer number of trials

conf.level

the confidence level for interval estimation; defaults to 0.95

Value

• Returns an object of type prop.ci, which includes:
• x
  the number of successes
• n
  the number of trials
• conf.level
  the confidence interval level
• lower
  the lower bound of the confidence interval
• upper
  the upper bound of the confidence interval
• interval
  the interval
• estimate
  the estimated proportion ($(x+2)/(n+4))$

References

Agresti, A. and B.A. Coull. 1998. Approximate is better than "exact" for interval estimation of binomial proportions. The American Statistician 52: 119-126. http://www.jstor.org/stable/2685469

See Also

binom.test

Examples

propCI(7, 50)
propCI(7, 50, conf.level = 0.99)
# should be very close to the score interval of prop.test
prop.test(7,50)