Statistics and risk

Statistical power

The ability of a study to detect an effect of a specified size if it truly exists.

What it means

Definition

Statistical power. Statistical power depends on sample size, effect size, event rate and variability. Underpowered studies can miss real effects or produce exaggerated positive estimates when results happen to be significant.

What it can tell you

It can help clarify what sort of evidence or claim is being discussed.

What it cannot tell you

It cannot, by itself, prove that the headline is accurate or clinically important.

Headline trap

Why this matters in headlines

A negative small trial may be inconclusive rather than proof that an intervention does not work.