VetsLearn — Analyze data & support decisions — 06: Capstone: produce a reproducible decision analysis
Revision 2026-10-03; fictional training fixture.

OUTCOME
Deliver a calculation package a colleague can inspect.

INDEPENDENT TASK
Independent CSV-style fixture: id,method,minutes,pass. M1,manual,10,yes; M2,manual,12,yes; A1,assisted,6,yes; A1,assisted,6,yes; A2,assisted,8,no; A3,assisted,,yes. IDs are unique cases; duplicate A1 is accidental. A2 repair took 7 additional minutes. Workloads are not matched.

STEPS
1. Preserve all original rows and log duplicate/missing entries.
2. Compute known effort and pass rates over distinct cases.
3. Include A2 rework; keep A3 effort unknown.
4. Write a 200-word recommendation and a matched follow-up test plan.

ACCEPTANCE CRITERIA
- Manual known effort 22 minutes; assisted known corrected effort 21 plus unknown A3.
- Initial assisted pass rate 2/3 after deduplication; manual 2/2.
- Do not infer overall savings from incomplete unmatched data.
- Keep formulas, exclusions and uncertainty inspectable.

EXPLAIN YOUR APPROACH
Which missing value most affects your decision, and how would you collect it?

Keep your artifact, corrections, evidence and limitations yourself.
Self-review is not independent assessment. No paid model calls are required.
Review notes: /paths/analysis/06/#practice