M3: Project Update
The M3 submission in the NHANES worked example shows one illustrative version of this milestone. Use it for structure and depth — your project must be your own.
Purpose
M3 turns the preliminary analysis into a project draft that is ready for feedback. The update should show the current evidence, the intended KT product, and the main decisions that still need peer or instructor critique.
Due Timing
Submit by the Week 10 deadline posted in Canvas. This milestone supports report drafting, publishing checks, and the Week 11 reproducibility review.
Points and Score Ownership
M3 is worth 5 course points and receives one group score. The checklist below guides the direct 0–5 score; no contribution multiplier or performance-band conversion is used.
The five points are allocated across these broad criteria:
| Criterion | Points |
|---|---|
| Project update framing, current finding, and decision needing review | 1 |
| Rendered, evidence-based report draft | 1 |
| Dashboard/KT preview, audience fit, and caution | 1 |
| Targeted feedback request and reproducibility guidance | 1 |
| AI-use audit and complete, synced submission evidence | 1 |
| Total | 5 |
These are broad judgments, not collections of micro-deductions. M3 is a draft checkpoint: substantively complete work can earn full credit while still identifying choices and revisions for later stages.
Required Deliverables
Create or update:
Include these files. Seven files looks like a lot, but only report-draft.qmd involves substantial work, and it grows out of your M2 analysis; the .md files are short notes. Tags follow How To Read Submission Lists.
project-update.md[required]report-draft.qmd[required]report-draft.html[required]dashboard-preview.md[required] [brief — 1-3 focused bullets or sentences are enough]reproducibility-note.md[required]feedback-request.md[required]ai-use-note.md[required]
Commit report-draft.html with its source. The course-wide rule is that every rendered artifact explicitly listed by an assessment is part of the submission.
What To Submit
Submit your GitHub repository link on Canvas after committing and syncing the files above. In the Canvas comment box, include the latest commit hash, the folder path, and a rendered report link [optional — include only if relevant to your project]. A link only exists if you published or shared a rendered preview; the committed report-draft.html is enough on its own.
Update Requirements
project-update.md should include:
- current project title and audience
- refined research or KT question
- data source and analytic sample
- one finding that appears stable
- one finding or choice that needs review
- next steps before peer review
dashboard-preview.md is a lightweight preview file. Include a link, screenshot filename, or one short note describing where the current dashboard/KT draft can be found. It is not a full report.
dashboard-preview.md should include:
- link or screenshot path for the dashboard-style product
- intended audience
- one interaction, filter, or decision-support feature
- one limitation or privacy caution
feedback-request.md should ask for targeted feedback on:
- correctness
- interpretation
- visual clarity
- reproducibility
- audience fit
Reproducibility Expectations
The report draft must render in GitHub Codespaces or a documented local environment. Relative paths are required. Generated figures and tables should come from code, not manual copying.
AI-Use Expectations
AI may help revise prose, identify unclear sections, or suggest checks. Students must verify citations, numeric results, interpretations, and any code suggested by AI. The ai-use-note.md should name the main AI-assisted tasks and the verification steps.
Full-Credit Preparation Checklist
A substantively complete submission should show:
- report draft renders
- current findings are supported by code-generated evidence
- dashboard-style product is visible or clearly previewed
- feedback request is specific
- reproducibility note explains how to rerun the work
- AI use is disclosed and audited
Definition Of Done
M3 is done when another student can open the repo, render the draft report, review the dashboard preview, and provide useful feedback without guessing what the project is trying to do.