School Accreditation Check · method and verification

How was the School Accreditation Check checked, and what did the independent check find?

A separate reader recomputed the figures from the raw files and matched them with no mismatches. The method was written and hashed on 12 November 2026 before any figure was published, and 1 dated amendment records what the checking found.

What was fixed before any figure was published?

The method was frozen on 12 November 2026. The file's SHA-256 hash begins 2b39f09471ad, and the original, before the amendment, began c26ea7ec0614. The tool refuses to build if the method file no longer matches its hash.

What does the independent check do?

  1. Every sanction action in scope lands in exactly one episode (checked by the builder; the run stops otherwise).
  2. A separate reader recomputes, for a random sample of 200 institutions, the number of sanction actions and episodes straight from the raw action table, and the page's counts must equal them.
  3. A school search that matches a name must return the same institution record the page shows.
  4. A school whose institutional record is ended, denied or closed is never shown as currently accredited.

What did it find?

MeasureResult
Institutions sampled400
Seed99
Mismatches0
All institution episodes independent3,258
All institution episodes builder3,255

Sampled institutions match exactly. The whole-population difference is 3 institutions where two actions share a date and a sequence number, so their order is undefined in the data (protocol amendment 1).

The checking led to one amendment to the method, dated and recorded:

2026-11-12 (amendment 1, before any publication of the episode counts as a dataset): the control in section 6.2 was run. A separate reader that sorts each agency's actions by date then DAPIP SequentialId, with no shared tables with the builder, matched the builder exactly on random samples of 200 and 400 institutions (0 mismatches). Across all institutions it counts 3,258 episodes against the builder's 3,255. The difference is 3 institutions where a removal and a new sanction (or a sanction and a terminal action) carry the same date and the same SequentialId, so their order is not defined by the data; the builder and the reader order such ties differently. The published count is the builder's, 3,255, and the limit is stated in section 7: where two actions share a date and a sequence number, the order of episodes can differ by one. The original protocol hash is kept in PROTOCOL-SCHOOL-WARNINGS-2026-11-12.original.sha256.

What is still not covered?

DAPIP depends on what accreditors report and when; an action may appear after its date. Removal actions are recorded inconsistently across agencies and years, so "no removal recorded" is common and does not mean a sanction is still in force. Many episodes are heightened monitoring or focused review, which several accreditors use for ordinary follow-up. Some institutions appear under more than one identifier. Programmatic accreditation and Heightened Cash Monitoring are separate records.

How can you check it yourself?

The source data is public: see the Department of Education accreditation database. The dataset, the method and the code that rebuilds it are published at doi:10.5281/zenodo.24172646. Open the School Accreditation Check to see the result.

Questions about how the check works

What does "frozen before the first publication" mean?

The method was written, hashed and saved before any figure from the data was published. A change after that is an amendment with a date, and the original hash is kept, so anyone can see what was decided and when.

Why does an independent reader matter?

A tool can agree with itself and still be wrong. A separate reader, written without the builder's code, recomputes the figures from the raw files; where the two disagree, one of them has a defect.

Does a clean check mean the tool is always right?

No. It means the numbers match an independent recomputation under the stated method. The limits section lists what the method does not cover.

Further reading: Zenodo, where the datasets are deposited; the FAIR data principles; the National Academies on reproducibility; the author's ORCID record; the Department of Education accreditation database.

Compiled by Vincent Wesley Couey (ORCID 0009-0005-6869-308X).

☆ Save