What survives
The moon may affect sleep slightly, most plausibly through light, and the evidence is genuinely unsettled — which is why this volume is rated mixed rather than disconfirmed. Separately and more usefully: a visible, unmissable twenty-nine-day cycle is an excellent scheduling peg for a recurring review, and that works regardless of the physics.
What does not
Lunar effects on behaviour, birth rate, psychiatric crisis, crime and mood. This has been tested more thoroughly than almost anything else in the series and the ceiling is about one per cent of variance. The folk belief persists because memorable nights get attributed to the moon and unmemorable ones do not.
Rotton and Kelly (1985)
Psychological Bulletin 97(2), 286-306 · 37 studies reviewed
Tested Meta-analysis of 37 published and unpublished studies relating lunar phase to psychiatric admissions, crisis calls, homicide and other offences
Found Lunar phase accounted for no more than one per cent of variance in any outcome. The authors concluded that further research was unnecessary.
Since Yes; subsequent large datasets have not disturbed it
Cajochen, Altanay-Ekici et al. (2013)
Current Biology 23(15) · 33 participants
Tested Whether lunar phase affects objectively measured sleep in a controlled laboratory
Found Around the full moon, deep-sleep EEG activity fell about 30 per cent, sleep latency rose by five minutes and total sleep fell by twenty
Since Contested. Larger re-analyses, including Cordi et al. 2014 with over 1,200 participants, found no effect. The original sample was small and the analysis was not pre-registered.
Casiraghi, Spiousas et al. (2021)
Science Advances 7(5)
Tested Sleep timing across lunar cycles in three Argentine indigenous communities and US students
Found Sleep onset shifted later and duration shortened in the nights before a full moon, in communities with and without electricity
Since Recent and not yet independently replicated; the proposed mechanism is nocturnal light, not lunar influence
What would change this rating
A large pre-registered replication of the sleep finding. Cajochen suggested it, Cordi's larger sample contradicted it, and Casiraghi revived it with a light-based mechanism. This is a live question and the volume treats it as one.
A rating that no possible finding could
alter is an opinion. Every volume names the result that
would move it.