Detectable paired change under a declared planning model

Detectable paired change under a declared planning model

Usage

detectable_change(n, sd_change, design_effect = 1, power = 0.8, alpha = 0.05)

Arguments

n

Number of paired observations.

sd_change

Assumed standard deviation of individual changes.

design_effect

Variance inflation relative to independent observations.

power

Desired power; default 0.8.

alpha

Two-sided test size; default 0.05.

Value

Normal-approximation standard error, interval half width and minimum detectable change, in the same units as sd_change. This is a planning approximation, not achieved power inferred from an observed effect.

Examples

detectable_change(n = 100, sd_change = .2)
#>     n sd_change design_effect power alpha   se half_width detectable_change
#> 1 100       0.2             1   0.8  0.05 0.02 0.03919928         0.0560317