📋 Teacher Cheat Sheet — Session 7: When Data Tricks You

Data Science for Young Minds · Grade 3 · Ages 8–9

~60 min Ages 8–9 Session 7 of 8 ND-Friendly

⏱ Session Agenda

Time	Block	What's Happening
0–5	🔁 Warm-Up	"Can data lie?" — quick show of hands + partner discussion
5–18	📖 Lesson 1–2	3 ways data can mislead: truncated axis, tiny sample, biased question
18–38	🎮 Activity	Myth Buster Cards — groups examine 4 misleading charts/claims and identify the trick
38–50	📖 Lesson 3–4	How to be a critical data reader · The 3-question test for any chart
50–56	✍️ Activity	Students apply the 3-question test to 2 charts on their worksheet
56–58	🔁 Recap	"What's the first thing you check when you see a chart?"
58–60	👋 Close	Preview Session 8: "Your turn to design a full data project!"

Key insight to land: Data doesn't lie — but people can present it in ways that mislead. A smart data reader checks the scale, checks the sample, and checks the question before trusting any chart.

📦 Materials Needed

Pencils Student worksheets Myth Buster Cards (printed or on board — see below) Optional: examples from real advertisements or news

💡 Myth Buster Cards work best as printed half-sheets — one misleading chart per card with space for students to write the trick they spotted. Alternatively, project them one at a time for whole-class discussion.

📚 Key Vocabulary

Misleading — giving a false or exaggerated impression, even if technically true

Truncated axis — a graph axis that doesn't start at zero, making differences look bigger

Small sample — asking too few people to make a reliable conclusion

Biased question — a question that pushes toward a certain answer (revisit from Session 2)

Critical reader — someone who questions what they see instead of accepting it immediately

Reliable data — data that is trustworthy because it was collected fairly and accurately

💬 Discussion Questions + Teacher Notes

"Can a chart be 100% accurate but still misleading?"
→ Yes! A truncated axis shows real numbers — but the visual exaggerates differences. The numbers are true; the picture lies. This is the hardest concept in the session.
"If a cereal box says '9 out of 10 kids prefer this brand!' — what questions should you ask?"
→ Which 10 kids? Did they choose from many brands or just two? Were the kids selected fairly? Who paid for the survey? This connects to bias and sample size.
"Why would someone WANT to make a misleading chart?"
→ To sell something, to win an argument, to make their team look better. Emphasize: even good people make these mistakes accidentally. Always check — it's not always intentional.
"How can you tell if a scale is truncated?"
→ Check the y-axis — does it start at 0? If the bottom is cut off (starts at 50, 80, etc.), bars look huge even for small differences. A broken-axis symbol (≈) sometimes marks it honestly.
"What would make a chart you DON'T trust become trustworthy?"
→ Fix the scale to start at 0; increase the sample; use a fair question; show the raw data. Good habit: always ask "what would make this MORE reliable?"

🎮 Myth Buster Cards — 4 Scenarios

Present each card. Groups identify the trick and explain how to fix it.

Card 1 — The Tiny Jump

A bar chart shows Brand A scored 82 and Brand B scored 80. The y-axis starts at 78. Brand A's bar looks 5x taller than Brand B's. Trick: truncated axis.

Card 2 — Three Friends

"We asked 3 people — 2 prefer our app! That's 67% of users!" Trick: tiny sample (3 people is not enough to represent all users).

Card 3 — Leading Question

Survey: "Don't you think our new playground is amazing? Yes / Kind of." No "No" option. Trick: biased question with no negative option.

Card 4 — The Missing Category

A chart shows 3 categories for "favorite color" but the colors shown are only red, blue, green — and many students said "purple" which isn't shown. Trick: incomplete categories hide real data.

🔍 The 3-Question Test

Post this. Students apply it to any chart:

Does the scale start at 0? If not — the bars may be exaggerated.
Was the sample large and fair? If only a few people were asked, or only certain people — results may not represent everyone.
Was the question fair (not leading)? If the question pushed toward one answer, the data may be biased.

Pass all 3? The data is probably trustworthy. Fail any? Look more carefully.

✍️ Wrap-Up Prompt

Write on board:
"If you saw a chart in an ad that said '8 out of 10 dentists recommend this toothpaste' — what questions would you ask before believing it?"

5 min. Bridge to Session 8: "You've learned every part of the data cycle — now you'll use ALL of it in your own project!"

🧠 ND-Friendly Tips

Frame positively — This session can feel like "everything is a lie." Reframe: "Smart readers ask questions. YOU are becoming a smart reader."
Real examples — Students who struggle with abstraction benefit from real ads or news headlines. Cereal boxes and sports stats work great.
Myth Buster Cards — Work in small groups, not individually. Spotting the trick is more accessible when discussing aloud with peers.
3-Question Test — Give students a laminated card with the 3 questions. Some will use it for the rest of the year.
Normalize skepticism — Some students are reluctant to question "official" sources. Validate: "Even scientists check other scientists' work. That's how we make sure data is reliable."