Wilson Score Confidence Intervals

"This function computes a confidence interval for a proportion. It is based on inverting the large-sample normal score test for the proportion." (Alan Agresti, who wrote the original R code)

Inputs for success, total, and level are vectorized.

Usage

scoreci(success, total, level = 0.95, return_midpoint = FALSE)

Arguments

success

Success count.

total

Total count.

level

Optional. Confidence level. Defaults to 0.95.

return_midpoint

Optional. Return midpoint of confidence interval? Defaults to FALSE.

Value

Data frame:

• success Success count

• total Total count

• estimate Proportion

• conf.low Lower bound of the confidence interval.

• conf.high Upper bound of the confidence interval.

• midpoint Mid-point of the confidence interval (for return_midpoint = TRUE).

• level Confidence level.

See also

http://users.stat.ufl.edu/~aa/cda/R/one-sample/R1/index.html

Agresti A, Coull BA. Approximate is better than "exact" for interval estimation of binomial proportions, Am Stat 1998;52:119-126.

Brown LD, Cai TT, DasGupta A. Interval estimation for a binomial proportion (with discussion), Stat Sci 2001;16:101-133.

Examples

scoreci(success = 5, total = 10)
#>   success total estimate  conf.low conf.high level
#> 1       5    10      0.5 0.2365931 0.7634069  0.95
scoreci(success = c(5:10), total = 10, level = 0.9)
#>   success total estimate  conf.low conf.high level
#> 1       5    10      0.5 0.2692718 0.7307282   0.9
#> 2       6    10      0.6 0.3516386 0.8057730   0.9
#> 3       7    10      0.7 0.4416998 0.8731234   0.9
#> 4       8    10      0.8 0.5407928 0.9314420   0.9
#> 5       9    10      0.9 0.6522813 0.9773651   0.9
#> 6      10    10      1.0 0.7870580 1.0000000   0.9