Findings
Everything this auditor has flagged, across all its versions, and what happened next: still open, applied in a later version, acknowledged, or rejected as a false alarm. A finding raised again by a later run is superseded by the newer one, so each problem is counted once.
143 findings: 118 open, 24 applied, 1 rejected. 3 repeats were superseded by a later finding.
-
Version 4 gives each figure the same rounding everywhere: 114.6%, $33.9 billion, $72.7 billion, $51.5 billion, $21.2 billion and 52%. It spells out LEAP (Legislative Evaluation and Accountability Program Committee), replaces FTE with "full-time-equivalent", and explains the McCleary levy swap: a 2017 law raised the state school property tax and capped local levies from 2019. Fiscal years are written as FY2014, FY2019 and FY2025, with the term defined at first use.
-
Version 4 removes the hand-written Independence paragraph from the editorial note; the page generates that note. The summary is now one sentence naming the strongest lines of evidence: the Legislature's records (general fund and all funds), NASBO's state-reported data, and the population, price and income benchmarks.
-
In version 4 the answer reads: "Washington's general-fund appropriations rose 114.6% from 2013-15 to 2023-25, far outpacing population and inflation." It names the state, the budget and the period, so it stands alone when quoted.
-
In version 4 the seoTitle states the finding: "Washington's state budget more than doubled in a decade: fact check".
-
Version 3 uses $71.3 million and $228.7 million everywhere and "4th-most" throughout, spells out the U.S. Department of Housing and Urban Development (HUD) in the description and context, and explains respondent-driven sampling, rapid rehousing and the 95% prediction interval where they first appear.
-
Version 4 changes "programme" to "program" and spells out ppm, NTP, CDC, ADA, NHANES and FDA at first use. It explains dental fluorosis and network meta-analysis at first mention.
-
The version 4 summary is one sentence naming the Cochrane trial evidence and the NTP total-intake finding. The dosing guidance moved to key points.
-
Version 4 rates only the main IQ point ("Fluoride in toothpaste is a neurotoxin that lowers children's IQ"). The context covers the other parts, and the answer now reflects the open low-dose question.
-
The version 4 seoTitle states the finding: "Does Fluoride Toothpaste Lower Kids' IQ? Studies Concern Total Intake, Not Toothpaste".
-
EPA, TSCA and NTP are now spelled out at first use, and the article uses "Proposition 65" consistently.
-
The hazard title now says "possible". IARC, P&G and ppm are spelled out at first use, surfactants and ethoxylation are explained in the context, and No Significant Risk Level is capitalized the same way throughout.
-
Both uses of "confirms" now say "indicates".
-
The summary now names EPA's 2020 and 2024 exposure reviews and the product testing. No separate independence note was added: the article page already generates one from the evidence fields (4 methods, 7 institutions, 3 decades), and a hand-written note would duplicate it.
-
The title now states the positive finding ("EPA found no unreasonable cancer risk from using laundry detergents like Tide."), and the search title answers its question.
-
Spell out CSIRO (Commonwealth Scientific and Industrial Research Organisation) at first use.
CSIRO is now spelled out as Commonwealth Scientific and Industrial Research Organisation in the evidence institutions and the source name.
-
Added LROC NAC frame M175124932R (NASA SVS source), a link to APOLLO normal-point data, and the Chandrayaan-2 OHRC frame ID in ISRO's PRADAN archive.
-
The imagery limitation now states the resolution limits and that orbital images show hardware is present, not when or by whom it was placed; it also notes which imagery is NASA-only.
-
The conclusion now weighs the robotic-hardware alternative and states what still rests on NASA alone; context reduced to scope framing; editorial note adds caveats, an independence count (8 countries, 4 methods, 1969-2024), and review status.
-
Change "contradicted" to "False" in the conclusion and description to match the current label.
The conclusion now states the claim is false, and the description now says the evidence shows Apollo 11 put astronauts on the Moon.
-
Version 3 spells out SMART-1 (Small Missions for Advanced Research in Technology-1), SIR and NSSDCA at first use; briefly explains basalt, anorthosite, breccia, regolith, maria, differentiated, olivine and pyroxene, and orbital spectroscopy; and changes the image caption date to 2024-08-19 to match the ISO dates used elsewhere.
-
Version 3's geophysical-structure limitation now states that the item relies on a single NASA synthesis rather than independent interior studies, and that this synthesis shares Apollo sample, seismometer, and mass data with other evidence items.
-
Version 3's summary is one sentence naming the three most independent lines of evidence (returned samples, orbital infrared measurements, bulk density). The conclusion now weighs each line's independence and states what remains uncertain: the exact size and composition of the core, and how far the sampled sites represent unvisited regions.
-
Version 3 removes the stale "contradicted with high confidence" wording: the conclusion now says the claim is False, with high confidence, and the summary carries evidence only. The answer now names the Moon: "Samples, density, and remote sensing show the Moon is rock with a small metal-rich core."
-
Make seoTitle state the finding, e.g. "Is the Moon Made of Cheese? No: Samples Show Rock and Metal".
Version 3's seoTitle now states the finding: "Is the Moon Made of Cheese? Samples Show It Is Rock and Metal". socialTitle was also set to match the title, as the rules require.