The right test follows from two questions: what type is the output (Y), and what is being compared (X)? Then the
answer is a confidence interval and a practical conclusion, not only a p-value. scripts/pick_test.py chooses
the test; scripts/run_test.py runs the common ones on CSV data using scripts/_stats.py for the distributions.
Trigger Conditions
Activate this skill when the user:
- Wants to know if two machines, suppliers, shifts or methods really differ
- Needs to verify a suspected root cause with data (DMAIC Analyze)
- Asks which statistical test to use, or how many samples are needed
- Has before/after data and wants to prove an improvement
Interpreting Minitab output instead: minitab-analysis-interpreter.
Workflow
1. QUESTION -> practical question; Y type; X type; paired?; mean or spread?
2. PICK -> python scripts/pick_test.py --y continuous --x 2-groups [--paired] [--not-normal]
3. SIZE -> enough samples for the difference that matters? (assumptions-and-power.md)
4. CHECK -> normality, equal variances, independence
5. RUN -> python scripts/run_test.py t2 data.csv --col diameter --group machine
6. DECIDE -> p-value AND confidence interval -> practical conclusionExample
2-sample Welch t: L1 mean 12.00921 (n 30), L2 mean 12.01062 (n 30); t = -2.817, df = 56.8
95% CI for L1 - L2: -0.002404547 to -0.0004061197
p = 0.006665 < alpha 0.05: significant - the means differ.L2 runs about 0.0004-0.0024 mm larger: statistically real, and worth an offset correction only if it matters against the 0.018 mm tolerance.
Roadmap table, the eight steps and reading p-values: test-roadmap.md. Assumption checks, type I/II errors, sample size tables and common mistakes: assumptions-and-power.md.
Output Format
Deliver:
- Question in practical and statistical form (H0, H1, alpha).
- Test choice and why; assumptions checked.
- Result - statistic, p-value, confidence interval / effect size.
- Conclusion in plain words and the decision it supports; power caveat if not significant.
Scripts
python scripts/pick_test.py --y continuous --x 2-groups --paired
python scripts/pick_test.py --y proportion --x k-groupsExit codes: 0 recommendation, 2 combination not covered.
python scripts/run_test.py t1 data.csv --col diameter --mu 12.009
python scripts/run_test.py t2 data.csv --col diameter --group machine
python scripts/run_test.py paired data.csv --a before --b after
python scripts/run_test.py var2 data.csv --col diameter --group machine
python scripts/run_test.py anova data.csv --col diameter --group machine
python scripts/run_test.py mannwhitney data.csv --col diameter --group machine
python scripts/run_test.py prop2 --x1 12 --n1 400 --x2 4 --n2 380
python scripts/run_test.py chi2 table.csvExit codes: 0 not significant, 1 significant at --alpha, 2 usage or data error.
Additional Resources
| File | Content |
|---|---|
| test-roadmap.md | Y/X roadmap with alternatives, hypothesis testing steps, reading p-values |
| assumptions-and-power.md | Assumptions and remedies, error types, sample size tables, mistakes |
8d-problem-solving-report-writer
Write and review 8D problem-solving reports for customer complaints and serious nonconformities (automotive and aerospace practice: Ford Global 8D, VDA, GM, Stellantis formats): D0 emergency response, D1 cross-functional team and champion, D2 problem description with 5W2H and is / is-not and no cause in the statement, D3 interim containment within 24 hours with verification and clean point, D4 verified occurrence and escape (non-detection) root causes plus systemic cause, D5 permanent corrective actions for each cause verified before implementation, D6 implementation, validation with production data and removal of containment, D7 PFMEA, control plan and read-across to similar products, D8 closure and recognition. Checks an 8D discipline by discipline for missing content, human-error root causes, missing escape cause or action, late containment and dates out of order.
acceptance-sampling-plan-selector
Select and evaluate acceptance sampling plans for incoming, in-process and final inspection: ANSI/ASQ Z1.4 / ISO 2859-1 single sampling under normal inspection (code letter from lot size and inspection level I, II, III, S-1 to S-4; sample size, Ac and Re from AQL, following the table's arrows to the correct plan; 100% when n reaches the lot size; AQLs above 10 per 100 units), operating characteristic curves (Pa at any quality, producer's quality level, indifference point, LTPD, AOQ and AOQL, binomial or hypergeometric), plans designed from producer and consumer risk, and c = 0 zero-acceptance plans. Covers AQL choice by defect class, switching rules between normal, tightened and reduced inspection, when sampling is not appropriate, and random sampling.
approved-supplier-list-manager
Manage and audit an approved supplier list (ASL) / external provider register for ISO 9001 8.4.1, IATF 16949 8.4 and AS9100 8.4.1: supplier status (approved, conditional, probation, pending, disqualified), scope of approval by commodity and process, risk class, certification and expiry, re-evaluation intervals by risk, conditions with end dates, single-source contingency, approver and customer-directed sources. Audits the list for expired or expiring certificates, overdue re-evaluations, conditional or probation status past its decision date, high-risk suppliers without the required certification, single-source suppliers without backup and missing approvers, and checks purchase orders for suppliers that are off-list, disqualified or pending, or buying outside the approved scope. Covers ERP blocks, PO requirements, off-list and emergency purchases.
apqp-phase-planner
Plan and track Advanced Product Quality Planning (AIAG APQP 3rd edition, IATF 16949) for a new or changed automotive part: back-schedule the five phases (plan and define, product design, process design, product and process validation, launch and feedback) from the start-of-production date by program complexity, with overlap, PPAP lead and safe-launch period; list the deliverables and gate for each phase (DFMEA, DVP&R, feasibility, process flow, PFMEA, control plans, MSA, Ppk, run at rate, PPAP, safe launch, lessons learned); adjust for build-to-print suppliers; flag compressed timelines and late or untracked deliverables against a status list. Covers gate reviews, red/yellow/green status reporting, run at rate and safe-launch exit criteria.
as9100-gap-analyzer
Run an AS9100D / EN 9100:2018 gap analysis for aviation, space and defence suppliers on the requirements added to ISO 9001: on-time delivery and conformity focus, management representative, ethics and human-factors awareness, operational planning and operational risk management, configuration management, product safety, counterfeit-part prevention, special requirements, critical items and key characteristics, external provider register and flow-down, FOD prevention, tooling and NC program control, special process validation, first article inspection (AS9102), traceability, in-service data and nonconforming product control. Generates and scores a worksheet, grades missing core controls as major, and checks invalid n/a claims. Covers AS9101/AS9104, OASIS, AS9102 FAI forms, counterfeit standards (AS5553, AS6174, AS6081), NADCAP and the planned IA9100 transition.
audit-report-writer
Write internal, supplier and pre-certification audit reports with the content ISO 19011:2018 6.5 expects: audit objectives, scope, criteria with revisions, dates, team and auditees, method and sampling statement, specific strengths, findings table and three-part finding details (requirement, evidence, statement), conclusion on the degree of fulfilment of criteria, diverging opinions, follow-up deadlines and distribution. Builds a Markdown report from a JSON audit record, generates the conclusion from the findings (major / minor / none) and flags reports missing required content or findings without clause or evidence. Covers conclusion wording, closing meetings, response expectations and verifying corrective actions.