# SA-MET-GRR-001 - Gauge R&R and Metrology-System Decision Replay

Version: 2026-08-23-v94

## Purpose

Teach how to review a measurement-system study without collapsing repeatability, reproducibility, bias, linearity, stability, resolution, uncertainty, representative part range, and study design into one pass/fail number.

## Deterministic sweep

4 study designs x 4 part-spread states x 4 repeatability states x 4 reproducibility states x 4 bias/stability postures x 4 review roles = 4096 rows.

## Calculation contract

- Direct deterministic variance-component proxies only; no raw observations are fitted.
- No ANOVA, REML, mixed-effects model, Minitab execution, or spreadsheet macro is run.
- The distinct-category value is an explicitly labeled proxy: floor(1.41 x part-to-part proxy / total Gauge R&R proxy), bounded below at one.
- No percent-tolerance value is published because no owned customer or process tolerance is present.
- Fixed bias, range linearity, stability/drift, and resolution remain separate from Gauge R&R.
- The combined uncertainty review proxy is not a GUM standard or expanded uncertainty.

## Gate coverage

- bounded_measurement_system_teaching_case: 12
- reference_bias_review: 16
- range_linearity_review: 16
- stability_drift_review: 16
- operator_reproducibility_review: 64
- tool_configuration_reproducibility_review: 64
- operator_part_interaction_review: 64
- repeatability_variation_review: 128
- repeatability_resolution_review: 128
- representative_part_range_required: 256
- part_family_confound_review: 256
- discrimination_review: 4
- nested_equivalence_evidence_required: 1024
- unbalanced_cells_repair_required: 1024
- confounded_design_rebuild_required: 1024

## Local source-review lineage

- `08_references/semiagora_simulation_atlas/DAILY_CONTENT_TRIAGE_2026-08-22_191627.md` - SHA-256 `EF092D57C8861E18EFC06A5DB801A5839314EF30C650960A71D5AC9C3B82ED4E`
- `08_references/semiagora_simulation_atlas/daily_content_triage_2026-08-22_191627.json` - SHA-256 `97DFA853F2970DFB98A34D4BAC2A96382ADAB2AB1E24962B8252FD30800360C7`
- `07_assets/factory_lab/daily/2026-08-22/20260822_191627/FACTORY_RUN.md` - SHA-256 `0728E23829728D3CD884C44CB71E17FB314E36BD50C2C7311D795C7973720F17`
- `07_assets/factory_lab/daily/2026-08-22/20260822_191627/resource_scan/daily_resource_candidates.csv` - SHA-256 `44AF6757325927E45C05449BA9C558E393C13250501DCBDD737AA5901098A6A3`

## Link-only source spine

- [NIST Gauge R&R overview](https://www.itl.nist.gov/div898/handbook/mpc/section4/mpc4.htm) - Official overview separating repeatability, reproducibility, stability, bias, resolution, linearity, drift, gauge differences, and uncertainty topics.
- [NIST Gauge R&R study design](https://www.itl.nist.gov/div898/handbook/mpc/section4/mpc42.htm) - Official design anchor for representative artifacts, operators, gauges, configurations, and the conditions included in a measurement study.
- [NIST repeatability analysis](https://itl.nist.gov/div898/handbook/mpc/section4/mpc441.htm) - Official repeatability vocabulary anchor. No NIST example values or tables are copied.
- [NIST reproducibility analysis](https://itl.nist.gov/div898/handbook/mpc/section4/mpc442.htm) - Official reproducibility vocabulary anchor. No NIST example values or tables are copied.
- [NIST uncertainty from a gauge study](https://www.itl.nist.gov/div898/handbook/mpc/section4/mpc46.htm) - Official uncertainty anchor. The SemiAgora combined uncertainty field remains a teaching review proxy, not a GUM uncertainty result.
- [NIST gauge linearity](https://www.itl.nist.gov/div898/handbook/mpc/section4/mpc452.htm) - Official anchor for checking changes across a measurement range while keeping linearity distinct from Gauge R&R.
- [NIST differences among gauges](https://itl.nist.gov/div898/handbook/mpc/section4/mpc454.htm) - Official anchor for between-gauge and configuration differences while preserving study-design context.
- [BIPM VIM measurement precision](https://jcgm.bipm.org/vim/en/2.15.html) - Official metrology vocabulary anchor for precision conditions and the distinction between repeatability and reproducibility conditions.
- [Minitab crossed and nested Gage R&R study forms](https://support.minitab.com/en-us/engage/help-and-how-to/tools/forms/types-of-forms/statistical-analysis/gage-r-r-study/) - Official vendor documentation used only to anchor crossed-versus-nested study vocabulary; no software output or example dataset is copied.
- [Minitab number of distinct categories](https://support.minitab.com/en-us/minitab/help-and-how-to/quality-and-process-improvement/measurement-system-analysis/supporting-topics/gage-r-r-and-wheeler-s-emp-studies/using-the-number-of-distinct-categories-in-a-gage-r-r-study/) - Official vendor documentation anchor for a distinct-category diagnostic. The public replay labels its value as a proxy, never an acceptance rule.

## Public boundary

- All 4,096 rows are original deterministic teaching proxies; zero real measurement rows, zero owned measurement files, and zero external numeric values are used.
- No customer specification, tolerance, production acceptance threshold, universal Gauge R&R pass/fail band, quality disposition, lot release, qualification, calibration claim, or traceability claim is present.
- Repeatability and reproducibility remain separate from bias, linearity, stability/drift, resolution, and uncertainty review.
- No live ANOVA, REML, mixed-effects model, Minitab execution, spreadsheet macro, solver, upload, metrology-system connection, APC/FDC action, or equipment control is enabled.
- The browser filters a saved CSV only. A real study needs owned raw repeats, stable representative parts or artifacts, declared crossed/nested design, randomization, gauge/operator/configuration identity, references, environmental conditions, analysis, and accountable review.
- No foundry, fab, process, product, customer, supplier, instrument, operator, method, or software performance claim is made.