How we score:
Verdicts come from random-effects meta-analyses with GRADE certainty
assessments where the data supports it; impact is sized in MID-units
(|MD| / instrument MCID) when a native measurement scale exists, and
falls back to Cohen's d otherwise. We
report null findings prominently and don't hide mixed evidence.
Read the full methodology →