engineering.uncertainty.coverage_factor

Three distinct suites: educational examples on the calculator page, this legacy regression count (10/10 golden + boundary), and the CVP verification suite below. They are not the same number. Property, metamorphic, round-trip, O3, and interface statuses/checks are recorded in the CVP Manifest . SKIP (0/0) means that suite has no published vectors.

This calculator CURRENT · Public schema 1.0.0 matches · Semantic contract ✓ · Production attested · Public/cache ✓ · Origin ✓

CVP verification suite

Protocol checks from the Evidence Manifest. Distinct from the legacy regression suite above. Verification revision 2026-09-16.o2-o3. Calculation version 1.0.0. CVP 1.0.0-proposed. 5/5 golden · 5/5 CVP boundary · 7/7 invalid · 3/3 property · 2/2 metamorphic · 8/8 O3 · 4/4 cross-interface · 4/4 CVP contract. digest 22b395d8c5f7. O2: two-sided k = t_{(1+p)/2}(ν) vs separate-module inverse (|Δk|≤2e-6). Infinite ν uses Φ⁻¹. Not U=k·u_c, not one-sided k. ≤2e-6 vs O3 applies only to the published tabulated k vectors (ν=8/1/2/30, inf, default p, p=0.99, nu_eff alias). It is not a whole-domain guarantee and does not cover U=k·u_c or one-sided k. Infinite-ν comparison is limited by the IUT Hastings erf polyfill when Math.erfc is absent.

• Golden (oracle-backed) 5 / 5
• CVP boundary 5 / 5
• Invalid input 7 / 7
• Property 3 / 3
• Metamorphic 2 / 2
• O3 high-precision 8 / 8
• Cross-interface 4 / 4
• CVP contract 4 / 4
• Interfaces PASS · UI (SSR) / REST / MCP
• Interface claim ui-ssr is query-result HTML, not a live browser session. Error-path engine·REST·MCP 1/1 (status, code, calculation_version). SSR compared on URL-canonical requested calculations; empty query is idle (not an error) and JSON-typed object/array inputs are REST/MCP-only.
• Verified scope Two-sided k = t_{(1+p)/2}(ν), including ν→∞ → Φ⁻¹((1+p)/2), + O3 mpmath tabulated k.
• Assurance Engineering
• Deferred Not expanded U = k·u_c, not one-sided k, and not Welch–Satterthwaite ν_eff.

• release artifacts aligned PASS
• schema valid input executes PASS
• engine success output validates schema PASS
• schema rejects incomplete input PASS

Fingerprint sha256:52484b2831509e6e36865a8b34a6238ce26d258becea8369ba2a0d004ef9bebc

Input schema sha256:20893f1fb226455a702cbda171b92f0b4b3d1c226266deecc18af0c59cd6ec83 · Output schema sha256:daaf8adbcbffabcd2ab99d1546bea58f7b62ec619902a8a0141e9ea3698fd2a1 · Build-time SHA-256 digests prove published artifacts agree with each other (release_integrity.scope=workspace). Production attestation GETs canonical URLs twice (public/cached view and origin), including Capability HTML, via npm run attest:production --write (published at /.well-known/calculatorx-production-attestation.json) and fails if the public Capability page advertises a different calculation_version. This is artifact consistency, not an independent immutable release log.

Tool

coverage-factor

Engine

coverage-factor

Engine build

2026-09-16

Evidence core

2026-09-16T05:28:40.984Z

Rendered

2026-09-16T22:03:43.735Z

Last reviewed

—

Assumptions

• JCGM 100:2008 G.3 / G.4 two-sided coverage factor
• Default confidence p = 0.95
• Infinite ν uses the standard-normal quantile
• Not expanded U = k·u_c and not GUM model combination

Limitations

• Does not compute expanded U = k·u_c (use Uncertainty propagate)
• Does not combine a measurement model
• Does not look up ν_eff (use Welch–Satterthwaite)