🎓 Instructor Command Centre

60-minute facilitation guide - Last Class

This page is for Dr. B eyes. Students can read it too as they might want to go over the materials on their on after the semester is over. That said the flow is optimised for Dr. B facilitation.

60-minute plan

Time	File	What you do	Key talking point
0:00–0:05	index.qmd	Open story, read hook aloud	Set the quest frame; this is class-wide
0:05–0:18	demo.qmd	Live narrated demo	YAML → chunks → inline R → callouts
0:18–0:33	report.qmd	Walk through EDA + tidymodels sections	Point at reproducibility at each step
0:33–0:43	slides.qmd	Render and present the deck	Show Space key navigation; swap a theme live
0:43–0:55	missions.qmd	Class reflection + debrief prompts	Collect all four fragments together
0:55–1:00	Final reveal	Read the toolkit definition aloud	Close the quest narrative

---

Before-class checklist

• Clone / unzip the project folder and open as RStudio project
• Run quarto::quarto_render() to pre-render the whole site
• Open all rendered HTML files in browser tabs — have them ready
• Optionally run quarto::quarto_preview() for live reload during demo
• Tell students at the top: “You do not need R running. Watch the screen and engage.”
• Required packages: tidyverse, tidymodels, GGally, patchwork

install.packages(c("tidyverse", "tidymodels", "GGally", "patchwork"))

---

Segment-by-segment notes

Stage 0 · RStudio Intro (10 min)

Show what RStudio looks like and how to navigate it.

Stage 1 · demo.qmd (13 min)

Have source open left, HTML preview right. Narrate the render loop.

Key moments: 1. YAML — swap cosmo → darkly live. Students gasp every time. 2. Inline R — “The sentence updates when the data updates. No copy-paste.” 3. code-fold: show — click the code triangle. Show them the toggle. 4. Callout types — ask students which type they would use for a model limitation disclaimer.

---

Stage 2 · report.qmd — EDA section (10 min)

Key moments: 1. glimpse() and other exploration functions — “This replaces opening Excel to ‘get a feel’ for the data.” 2. Pair plot — point at the correlation between clues_solved and final_treasure. [not covered in class but useful in real life] 3. Box plots by major — “Which major would you want to hire based only on this chart?”

---

Stage 3 · report.qmd — Tidymodels section (10 min)

Key moments: 1. The workflow pipeline — draw split → recipe → spec → workflow → fit → predict → evaluate on the whiteboard or annotate on screen. 2. RMSE — ask a student to explain RMSE in plain English before revealing the value. 3. Diagnostic charts — residuals vs fitted: “What pattern tells us the model is wrong?” Predicted vs actual: “Perfect model → dots on the diagonal.”

---

Stage 4 · slides.qmd (10 min)

Open rendered .html fullscreen.

Key moments: 1. Press F for fullscreen, Space to advance. 2. Show the source .qmd side by side for 30 seconds — “This is all it takes.” 3. Change theme: simple → theme: moon and re-render live (~15 seconds). Students can suggest the theme. 4. Ask: “What slide would YOU add to describe your major’s relationship with data?”

---

Stage 5 · Final chamber (12 min)

Open missions.qmd. Read the four fragments aloud together.

Debrief prompts (pick 2–3): - “What would happen to a report you built last semester if the raw data file changed?” - “When would you choose a website over a PDF?” - “What is the business value of reproducibility in a professional setting?” - “Where in your career — internship, job, grad school — would this workflow matter most?”

Close by reading the Golden Data Toolkit definition aloud.

---