Updates & corrections

Read CouplesData release notes, source coverage limits and the corrections policy for published relationship statistics.

Reviewed October 8, 2026 · automated editorial checks

October 8, 2026 — Marital-status research

Expanded the separately versioned research cohort to 7,745 observations across the United States, all 50 states and D.C., plus four counties and four cities. Authenticated Data Commons imports now supply the five-year history; 300 latest status counts and denominators match original Census tables. Searchable place selection and original Census one-year sex breakdowns accompany every supported profile. Inspect research release 2026-10-08-r2.

October 1, 2026 — Initial release

Published 5,237 observations across ACS, CDC/NCHS, Eurostat and ONS. Added a scoped Pew survey explainer and the first reproducible state comparison. The expanded r2 release adds five ACS measures (1,040 observations) without changing any r1 observations: recent reported marriages and divorces, current-marriage duration, and same-sex couple-household shares. The r3 source-note review preserves all estimates and adds CDC revision flags, population-basis footnotes and the Oklahoma cross-table coverage caveat.

Inspect release 2026-10-01-r3.

Release history

  • 2026-10-01-r3 — 5,237 observations; Compared CDC state and national footnotes. Added source revision flags and original table notes, including Oklahoma state/national coverage mismatch in 2000–2003. Verified zero changes to values, MOEs, denominators or non-flag observation fields. No independent expert review.
  • 2026-10-01-r2 — 5,237 observations; Verified retained Census table universes and component cells for added measures; r2 adds 1,040 observations with no changes to r1 observations. Sampling MOEs and person/household distinctions retained. No independent expert review.
  • 2026-10-01-r1 — 4,197 observations; Nine risk-focused tests passed. Census national values, table universes, CDC reporting exclusions and ONS workbook cells were compared to retained source records. Not independent expert review.

Corrections

No post-publication corrections have been recorded for this release. Any future correction will identify the affected claim, original and corrected values, reason and replacement release.

Known coverage limits

  • ACS baseline is 2021–2024, one-year estimates.
  • CDC baseline ends in 2023; five states do not report divorce data for that year.
  • European coverage is limited to five selected countries, with gaps preserved.
  • Dating evidence describes fieldwork in July 2022, not current behavior.
  • No HCMST microdata, global UN library or personal risk estimates.