Runs a prospective (a-priori) power analysis to determine the sample size, the minimum detectable effect (MDE), or the achievable power for a planned statistical test BEFORE the study runs — pairing the choice of test, the expected effect size (from pilot, prior literature, or a smallest-effect-of- interest), and the α / β trade-off. Refuses to compute post-hoc / observed- power on a completed study (which is mathematically a deterministic function of the p-value and adds no information). Use whenever the user is designing an experiment, asks how many subjects are required, is writing a grant or pre-registration, has a non-significant result and is tempted to argue from o…
