Skip to contents

Converts a fs_dgm_scale object into the sampling standard deviation of the within-region difference-in-means estimator at trial size n.

Usage

fs_scale_se(scale, n, region = NULL, jensen = FALSE)

Arguments

scale

An object of class "fs_dgm_scale".

n

Integer. Trial size.

region

Character. Region name, matching a row of scale$regions. Default NULL returns every region.

jensen

Logical. If TRUE, inflate by the exact factor accounting for the random region size \(n_g \sim \mathrm{Bin}(n, P(g))\) and random arm split \(n_1 \mid n_g \sim \mathrm{Bin}(n_g, p)\), returning the unconditional standard deviation. Default FALSE, which returns the idealised (fixed-count) value.

Value

Named numeric vector of standard deviations.

See also