Combine p-values using Simes' Method

Aggregates a vector of p-values using Simes' method, which controls the family-wise error rate under the union null hypothesis. This is particularly useful when testing whether at least one of several related tests shows a significant effect.

Usage

combine_simes(pvals)

Arguments

pvals

A numeric vector of p-values (e.g., from L single-effect models for one predictor).

Value

A scalar: the Simes-combined p-value.

Details

The Simes p-value is computed as: $$\min_{i=1}^L \frac{L \cdot p_{(i)}}{i}$$, where \(p_{(i)}\) are the ordered p-values.

Examples

if (FALSE) { # \dontrun{
p <- c(0.01, 0.2, 0.04)
combine_simes(p)
} # }