Sprint 3 planning: analysis, integration, and communication (Improve phase)¶
How to use: the Scrum Master fills this in during Sprint 3 planning on Thu 2026-10-29. Candidate items become GitHub issues in milestone Sprint 3. This sprint carries the largest weight (25%) and feeds the final demonstration (20%). Plan the report and presentation work as first-class backlog items, not leftovers.
Dates: Thu 2026-10-29 to Wed 2026-11-18. Review and retro: Wed 2026-11-18. Check-ins: Thu 2026-11-05, Thu 2026-11-12 (also rehearsal 1). Product Owner: Leticia Aderhold. Scrum Master: Jerad Dunne. Developers/Researchers: all four.
Sprint goal¶
Example shape: "By Nov 18 the report answers RQ1 to RQ3 with validated evidence, every figure is regenerated by the pipeline, a tagged release candidate exists, and the primary results reproduce from the README."
Sprint goal:
Capacity¶
| Member | Available hours this sprint | Planned points | Notes |
|---|---|---|---|
| Leticia Aderhold | |||
| Jerad Dunne | |||
| Hina Kramer | |||
| Total | Sprint 2 velocity: ___ points |
Candidate backlog (rubric deliverables for Sprint 3)¶
| # | Item (rubric deliverable) | Acceptance criteria | Estimate | Owner | Issue # |
|---|---|---|---|---|---|
| S3-1 | Final or near-final results addressing the research question | Results section states the answer with numbers produced by the pipeline | #27 | ||
| S3-2 | Robustness checks, sensitivity analysis, or comparison across meaningful subgroups | Similarity threshold, severity cutoff, front-matter-valid subset, and top-copied exclusion checks, each as a script and a table (results/sensitivity_*.csv) |
#28 | ||
| S3-3 | Error analysis and manually inspected examples | Error taxonomy with counts, adjudicated rater disagreements, and at least five inspected examples in the report | #29 | ||
| S3-4 | Final figures and tables generated from the repository pipeline | make figures regenerates every figure in figures/; no hand-edited images |
#30 | ||
| S3-5 | Complete draft report with citations and a clear separation between findings and interpretation | All rubric sections drafted in report/draft.md; Results contains observations only; Discussion contains interpretation; each section reviewed by a member who did not write it |
#31 | ||
| S3-6 | Threat-to-validity section addressing dataset bias, measurement error, confounding, missing data, reproducibility, and generalizability | THREATS_TO_VALIDITY.md complete, including rater agreement and sample limits, and mirrored in the report |
#32 | ||
| S3-7 | Documented release or tagged version of the repository | Tag v0.9.0-rc1 with release notes listing commands and outputs |
#33 | ||
| S3-8 | Final demonstration plan and presentation materials | docs/PRESENTATION_PLAN.md filled with a speaker for every segment; slides in report/ or linked; rehearsal 1 done Nov 12 |
#34 | ||
| S3-9 | Retrospective identifying the group's most important process improvement | docs/sprints/sprint-3/retrospective.md section 9 completed |
#35 | ||
| S3-10 | README reproduction in a clean environment | A member who did not write the README reproduces the primary table from a fresh clone following only the README; time and problems recorded | |||
| S3-11 | Traceability audit | Every figure, table, and decision in the report links to an issue, PR, commit, or ADR (checklist in the review) | |||
| S3-12 | Related work section with real citations | At least six sources in report/references.bib, each checked against its DOI or arXiv page |
|||
| S3-13 | FMEA-style ranking of capability categories | Severity from the rule file, occurrence from validated prevalence, detection from validated recall; table in results/ |
Risks pulled from the FMEA register¶
| Risk | RPN | Mitigation this sprint | Owner |
|---|---|---|---|
| Results change late and figures go stale | Figures only through make figures; CI regenerates on PR |
||
| Report writing concentrated on one member (R06) | 147 | Report sections owned by different members (W8 in docs/workspace/WORK_SIGNUP.md); descope order in P-01 section 10 if time runs short |
Product Owner |
| A member is unavailable near the deadline (R05) | 108 | A backup for every report section and demo segment | Scrum Master |
| Demo fails live | Pre-recorded run and static figures as backup (see docs/PRESENTATION_PLAN.md) |
Improve phase exit criteria (DMAIC)¶
- The improvement to the analysis (robustness, error analysis) is implemented and its effect on the answer is stated.
- The process improvement with the largest measured effect this semester is named with before and after numbers.
- Every report figure and table is reproducible from a tagged commit.
- The threats-to-validity table has a status for every row.
- Control plan drafted for finalization (
docs/lean-six-sigma/CONTROL_PLAN.md): what is checked, by whom, and when, before submission.
Planning meeting notes¶
- Date: 2026-10-29. Facilitator: Jerad Dunne.
- Attendees: [ ] Leticia Aderhold [ ] Jerad Dunne [ ] Hina Kramer
- Items pulled into the sprint (issue numbers):
- Items explicitly deferred and why:
- Carry-over from Sprint 2: