# Verification checks (P4)

1. Clean re-run: deleted out/results.json, re-ran analysis.py, byte-identical to previous output (cmp) — PASS.
2. n_verified (4) + len(unverifiable_references) (1) = 5 = n_analysed = number of references — PASS.
3. n_verified equals count of references with all three checks true — PASS.
4. Every locator begins with the source id — PASS.
5. n_input_rows match wc -l (README 17; sources 134/58/107/66/91) and JSON item count 5 — PASS.
6. Extract SHA-256 values equal data/README.md values — PASS (P1).
7. Arithmetic cross-checks printed by the script: exp(-0.431)-1 = -35.01% vs printed -35.0%; -0.268/0.366 = -0.732; -0.268*0.068/0.047 = -0.388; -0.0962/0.261 = -0.3686 vs printed -0.3685 (rounding) — PASS.
8. Baseline (title+year only) flags R3 only; full check also flags only R3 (fails all three) — consistent.
9. Information rules: only extracts used; years compared to first-submission and reproduced-version years from Bibliographic record lines — PASS.