⏱ Session Agenda
| Time | Block | What's Happening |
|---|
| 0–5 | 🎯 Hook | "Ice cream sales and drowning rates both go up in summer. Does ice cream cause drowning?" (Take votes.) |
| 5–18 | 📖 Lesson 1–2 | Correlation defined · Causation defined · The difference · Confounding variables explained |
| 18–40 | 🔬 Activity | 4 Spurious Correlations — students explain the real relationship for each; then apply to serious examples |
| 40–50 | 📖 Lesson 3 | How to argue from evidence responsibly · Claim-Evidence-Reasoning framework · Real scientific examples |
| 50–56 | 🧠 Brain Break | Thumbs up/down: "Correlation or Causation?" rapid-fire examples |
| 56–60 | ✍️ Close | Students write one C-E-R argument about a topic they choose. Preview S7. |
Tone note: Funny examples lower anxiety. Students learn better when they feel safe being wrong. Start with ice cream/drowning; move to serious examples only after the concept is secure.
📦 Materials Needed
Worksheets (1 per student)
Spurious correlation cards (printed or projected)
Pencils
C-E-R framework posted on board
💡 C-E-R Framework to post: Claim (what you think is true) → Evidence (data that supports it) → Reasoning (why the evidence supports the claim)
📚 Key Vocabulary
Correlation — two variables tend to change together (both go up, or one up/one down)
Causation — one variable directly causes the change in another
Confounding variable — a hidden third factor that explains why two unrelated things correlate
Spurious correlation — a correlation that exists but has no real causal relationship
Evidence — data or observations that support a claim
Claim — a statement about what is true, based on evidence
🍦 The 4 Spurious Correlations — Full Instructor Guide
1. Ice cream sales & drowning rates
Both rise in summer, fall in winter. Real connection:
hot weather (confounding variable). People swim more when it's hot → more drownings. People eat more ice cream when it's hot. Ice cream doesn't cause drowning.
Detective question: "Is there something else that's high in summer that explains both?"
2. Nicolas Cage movies & pool drownings
A real (and famous) spurious correlation from the internet. Both rise and fall together over time — pure coincidence. Real connection: none. Confounding: random variation in yearly data over small samples.
Use this to show how real the problem can look with charts. "The graph looks convincing — but there's no possible mechanism."
3. Shoe size & reading ability (in children)
Older children have bigger feet AND read better. Real connection:
age (confounding variable). Age drives both. A 10-year-old reads better than a 6-year-old AND has bigger feet — not because of feet.
Good example for students: "Would putting bigger shoes on a 6-year-old help them read?" Obviously no — the real cause is age/development.
4. Countries with more TVs per capita have higher life expectancy
Real connection:
wealth (confounding variable). Wealthier countries have both better healthcare (→ longer life) AND more TVs. TVs don't cause longer life. Buying everyone a TV wouldn't help.
This is a more serious example — good bridge to medical/policy claims where the stakes are real.
Key teaching move: After each example, ask "What is the confounding variable?" Then: "How would we test whether this is truly causal?" (We'd need a controlled experiment — hold everything else constant and change only the variable in question.)
💬 Discussion Questions + Teacher Notes
- "Ice cream doesn't cause drowning — so what does?"
→ Hot weather (the confounding variable) causes both ice cream sales to rise AND more swimming (→ more drownings). This is the classic example. Make sure students can name the confounding variable, not just say "it's wrong."
- "If two things are correlated, can we be totally sure one doesn't cause the other?"
→ Not just from correlation alone. You need a controlled experiment (changing only one variable) to test causation. Real scientists use randomized controlled trials (RCTs) for exactly this reason.
- "How could a politician misuse a correlation?"
→ "Cities with more police have more crime — so police cause crime!" (In reality, high-crime cities hire more police.) Direction of causation can also be backwards. This is called reverse causation.
- "What would it take to prove that exercise CAUSES better grades?"
→ A controlled experiment: randomly assign students to exercise vs. no-exercise groups, keep everything else identical, measure grades. Just correlating exercise habits with grades doesn't prove causation because many confounders exist (family income, school quality, etc.).
📝 C-E-R Framework — Arguing from Evidence
Post this on the board and refer to it throughout the session. Grade 5 students should start using this structure for all data-based arguments.
C — CLAIM: Make a clear statement about what you think is true.
"Ice cream sales do not cause drowning."
E — EVIDENCE: Cite specific data that supports your claim.
"Both ice cream sales and drowning rates peak in July when temperatures are highest."
R — REASONING: Explain why the evidence supports the claim — name the mechanism.
"Hot weather causes both increased swimming activity (→ drowning risk) and increased ice cream consumption. The confounding variable is temperature, not ice cream."
Watch for: Claims without evidence ("I just know"), evidence without reasoning ("the data shows it"), and reasoning that assumes causation from correlation.
🎯 Opening Hook
Write on board:
"Ice cream sales and drowning rates both go up every summer. Does ice cream cause drowning?"
Take a vote: Yes / No / Complicated. Accept all answers without judgment.
→ Most students will correctly say No — but ask: "How do you know? What's actually going on?" This is the lesson: naming the confounding variable.
🧠 Brain Break
Thumbs up/down rapid-fire:
"Correlation or Causation?"
- Smoking → lung cancer (Causation — proven by controlled studies)
- Shoe size → reading level in children (Correlation — confounded by age)
- Vaccination → reduced disease rates (Causation)
- Countries with more chocolate consumption → more Nobel Prizes (Correlation — wealth confounds)
🧠 ND-Friendly Tips
- Funny first, serious second — Ice cream/drowning and Nicolas Cage/drowning reduce the stakes. Once students feel safe, the serious examples (medical, political) land better.
- Concrete question for each example — Always ask: "What is the confounding variable?" This gives students a consistent scaffold for every new example.
- C-E-R posted and visible — Students with working memory difficulties need the framework externally visible throughout the writing portion.
- Validate "it's confusing" — Correlation vs. causation errors are made by professional researchers. Acknowledge the difficulty explicitly.
- Peer argument sharing — Students who struggle with writing can share their C-E-R verbally before writing. Oral rehearsal supports written output.