Claim in the Record
OBSERVEDAdded to the Record 2026-09-14Withdrawn — historical record

PR admissions: selected-period comparison

The claim

PR admissions, monthly: averages differ across two selected periods

In the 40-month window ending Jun 2026, the scan selects 32 earlier observations averaging 37,365 and 8 later observations averaging 28,954. The later period starts in Nov 2025. This split is at the edge of the dates the scan permits; it does not establish when a change began. These selected-period averages are descriptive, not a significance test. Seasonal patterns and the choice of window can affect the comparison; no cause is identified.

Method

Descriptive two-segment mean comparison selected by maximizing |difference in means| / pooled standard deviation, reported at ratios of at least 1.5 with at least eight observations in each segment. Boundary-selected split; no change-date or statistical-significance inference.

Key sources and licences

Added to the Record2026-09-14
How it is usedWithdrawn — historical record
Where it appearsHistorical reading retained on this receipt. The comparison was held for inspection rather than foregrounded in a product; it is now withdrawn from the current admitted corpus.
What would change itA new comparison requires independent review and a new admission. The earlier comparison cannot transfer its authority to another window or a lower reporting threshold.
Corrections2026-09-14: selected-period comparisons do not establish change dates or statistical significance; calculations are unchanged.
Withdrawn2026-10-04
Withdrawal reasonThe refreshed July 2026 admissions record no longer reproduces this comparison under the unchanged reporting threshold. The historical reading is retained here and is excluded from the current admitted corpus. No new comparison is admitted.
Technical details
Record IDR17-break:pr_flow
Historical candidate IDcand:R17-break:pr_flow:2025-11-01
Detection ruleR17-break
Indicatorpr_flow

This Receipt names the public sources and method needed to reproduce the figures it covers. Corrections are dated and explained, with the earlier published record retained.