Part 8: Wrap-up

What we learned and where to go next

What We Built Today

Over the past 3 hours, you created a complete reproducible workflow:

flowchart TD
    A["Raw Data"] --> B["data-cleaning.qmd"]
    B --> C["data_clean.rds"]
    C --> D["analysis.qmd"]
    D --> E["Tables & Figures"]
    D --> F["model_fit.rds"]
    E --> G["index.qmd"]
    F --> G
    C --> G
    G --> H["Word / PDF / HTML"]

    style A fill:#ffcdd2
    style H fill:#c8e6c9

Every step is documented, every change is tracked, and the entire pipeline can be re-run at any time.

Before vs After

Before After
Manual copy-paste of numbers Code-generated tables and figures
“Which version is this?” Git history shows every change
Fear of making changes Confidence to experiment
Hours to redo analysis Minutes to re-run pipeline
“I can’t remember what I did” Everything documented in code

Key Takeaways

Three principles to remember

  1. Automate with code → Reproducibility comes naturally when you eliminate manual steps

    • Function and iteration via purrr
    • Script-based publication-ready tables and figures

  2. Version control everything → Never lose work, experiment freely, recover from mistakes

  3. Plain text is powerful → Git-friendly, AI-friendly, future-proof

What’s Next

See the Resources page for learning materials, real-world examples, and links to continue your reproducibility journey.

← Part 7: Collaboration | Return to Home