关于R语言pROC包ci.se函数置信区间计算公式的问询

I totally get how frustrating it is when you need precise method details for your research, but the official documentation doesn’t spell things out clearly. Since the pROC package docs don’t explicitly state the formula behind ci.se(), here are some reliable ways to get the answer:

• Dig into the source code: This is the most definitive approach. In R, run pROC:::ci.se (the triple colon lets you access non-exported internal functions). Reading through the code will show you exactly which statistical method is implemented—you might spot references to exact binomial intervals, Wilson score intervals, or one of the other methods you found in that sensitivity/specificity CI paper.
• Check the package’s vignettes and citations: pROC includes detailed vignettes that often link to the statistical literature it relies on. Run browseVignettes("pROC") in your R console to access these documents; you may find a citation that directly maps to the formula used by ci.se().
• Reach out to the package maintainers: If the source code and vignettes don’t clarify things, contact the authors via the package’s GitHub repository or CRAN page. They’re usually happy to explain the underlying methods for research-related questions.

As a quick context note: many of pROC’s confidence interval functions default to methods like the Wilson score interval (a common choice for proportion CIs) or exact binomial intervals, but verifying via the source code is the only way to be 100% sure for ci.se().

内容的提问来源于stack exchange，提问作者Álvaro García Ródenas