A study too small to reliably detect the effect it's looking for.
An underpowered study has insufficient sample size (or events) to give a high probability of detecting a true effect of the expected magnitude, meaning a non-significant result may reflect low statistical power rather than a genuine absence of effect. Underpowered findings are also prone to inflated or unstable effect-size estimates when a significant result does occur by chance. Clinicians should treat null results from underpowered studies cautiously and check whether a pre-study power or sample-size calculation was reported.
This guide was auto-drafted and is pending editorial review.