September 24, 2026 · We got it wrong
- What it said
- The guide's headline statistic was corrected to 8,666 participants on 2026-06-03 (see the record of that date), but nine other fields kept the old figure of 9,350: the opening hook, the key takeaway, two body paragraphs, the skeptic note, a citation capsule, the AI citation hint, the search description and a visual's description.
- What it says now
- Every field says 8,666, the sum of the verified participant counts of the six studies in the guide that report one. The seventh, Schoenfeld 2017, reports no total, which is what the 2026-06-03 record explains. The audio script never carried the figure.
- Why
- The 2026-06-03 fix changed the hero statistic and the cross-references but did not sweep the guide's prose and metadata for the same figure. Found on 2026-09-24 when every correction record was checked against its live page, the day the records went public.
- Corrected
- Built on it
22 shorts
- Found by
- Our own checks of the site.