Definition

Statistical power

The probability that a study will detect an effect of a given size if it really exists — conventionally targeted at 80%.

Power is the sensitivity of your study: with this sample size, this effect size, and this alpha level, how often would you find the effect when it is truly there? Underpowered studies mostly produce false negatives — and when they do "find" effects, those estimates are inflated (winner’s curse).

Power analysis before data collection solves for the needed sample size; reporting it is now standard in preregistrations and methods sections. Post-hoc power computed from the observed effect is uninformative — it restates the p-value.