
Where possible the numbers were read from the institution's own page: the AAA Foundation crash study, the CDC sleep page, the NIOSH shift work bulletins, the AASM position statement and the CDC Yellow Book. The paper links on the slates were checked against the Crossref record for title, first author, journal and year. Some figures come only from search summaries of a paper, and each slate says which.
How to check any sleep claim
Ask who ran the study and how many people were in it
Check whether it is lab data, survey data or a wearable estimate
Look for the comparison group: compared with what?
Separate association from cause, as in the screens and heart attack studies
Be wary of exact percentages quoted without a link to the paper
Known weak spots, kept honest: the National Sleep Foundation ranges were not re-read in the 2015 paper, the heart attack figures on the daylight saving slate come from search summaries, and several slates rank below the institution's own page on authority, because they summarise it.
"A good sleep number comes with a source, a sample and a limit.
"Rule of thumb