1 / 15
Session 7 of 8 · Grade 3 Data Science
🕵️

When Data Tricks You

Spotting misleading charts and becoming a critical data reader
~60 minutes Grade 3 · Ages 8–9 Data Cycle: Evaluate
Today's Plan

What We'll Do Today

The Big Idea: Data doesn't lie — but people can present it in ways that mislead. A smart data reader asks questions before trusting any chart.
Warm-Up · 5 min

Can Data Lie?

"Can a chart be completely accurate
and still be misleading?"
Think about this…
A cereal box says "9 out of 10 dentists recommend!"
A news headline says "Crime UP 100%!"
An app says "Rated #1 by users!"
What questions pop into your head?
Today you'll learn to ask:
Does the scale start at zero?
How many people were asked?
Was the question fair?
What's missing from this picture?
Lesson 1 — Trick #1

Trick #1: The Chopped-Off Scale

A truncated axis is when the y-axis doesn't start at 0. This makes small differences look huge.
❌ Misleading — axis starts at 78
Brand A
82
Brand B
80
Scale: 78 → 80 → 82 → 84 (Brand A looks MUCH better!)
✅ Honest — axis starts at 0
Brand A
82
Brand B
80
Scale: 0 → 20 → 40 → 60 → 80 → 100 (nearly identical!)
Always check: Does the y-axis start at 0? If not — the differences may be exaggerated.
Lesson 1 — Trick #2

Trick #2: The Tiny Sample

"We asked 3 people — 2 prefer our app!
That's 67% of all users!"
😵 Why this is a problem
3 people cannot represent millions of users
One person changing their answer flips the result
67% sounds impressive — but it's just 2 out of 3
Small samples are unreliable (we learned this in Session 2!)
✅ What to look for instead
How many total people were asked?
Were they randomly chosen?
Do they represent the full group?
The bigger and fairer the sample, the more you can trust the result
Lesson 1 — Trick #3

Trick #3: The Loaded Question

Remember from Session 2: a biased question pushes people toward one answer. If the data came from a bad question — the chart is untrustworthy no matter how pretty it looks.
❌ Loaded question examples
"Don't you agree our school lunch is amazing?"
"Rate our service: Good / Great / Excellent"
"Most people prefer X — do you?"
No "disagree" or "no" option offered
✅ Fair question features
Neutral wording — no pressure built in
All reasonable options included
Not more than one question at a time
"How would you rate the school lunch: poor / okay / good / great?"
Activity · 20 min

🎮 Myth Buster Cards!

Your Mission

You'll examine 4 misleading data claims or charts
For each one: name the trick, explain why it's misleading, and say how to fix it
Work with your group — discuss before writing
Be ready to share your best catch with the class
Tricks to watch for:
Truncated axis (scale doesn't start at 0)
Tiny sample (too few people asked)
Biased / leading question
Missing categories (important options left out)
Myth Busters — 4 Cards

What Tricks Did You Spot?

Card 1
Brand A bar is 5× taller than Brand B — but they scored 82 vs. 80
✅ Fix: Start the axis at 0 so both bars are nearly the same height
Card 2
"67% of users prefer our app!" — based on 3 people
✅ Fix: Ask a much larger, randomly selected group before making that claim
Card 3
Survey only offered "Good / Great / Excellent" — no negative options
✅ Fix: Include a full range of options including "poor" and "okay"
Card 4
Chart shows 3 favorite colors — but "purple" was the most common answer and isn't shown
✅ Fix: Include ALL categories — even ones that weren't expected
Brain Break · 2 min
🚨

Trick or Trustworthy?

The teacher reads a data claim. Show thumbs up for TRUSTWORTHY or thumbs down for TRICK!

"9 out of 10 students prefer this school" — based on asking 10 friends

"In our survey of 200 randomly chosen students, 68% preferred outdoor recess"

"100% of people we asked loved our pizza!" — they only asked the pizza maker

Lesson 3

The 3-Question Test

Apply this test to any chart or data claim you see:
Question 1
Does the scale start at 0?
If NO → the differences may look bigger than they really are
Question 2
Was the sample large enough and fairly chosen?
If NO → the results may not represent everyone
Question 3
Was the question fair — not leading or biased?
If NO → the data may be skewed toward one answer
Pass all 3? Probably trustworthy. Fail any? Look more carefully.
Activity · 6 min

✍️ Apply the 3-Question Test

Open your worksheet — Part 3

You'll see 2 data claims with charts
Apply the 3-question test to each one
Mark which questions pass (✅) and which fail (❌)
Write: Is this data trustworthy? Why or why not?
Remember: a chart can fail one test and still have some useful information. Being critical doesn't mean rejecting everything — it means reading carefully.
Lesson 4

You Are Now a Critical Data Reader

Before trusting a chart
Read the title
Check the scale
Look for sample size
Ask who made it and why
Questions to ask
"Who was asked?"
"How many were asked?"
"What question was asked?"
"Is anything missing?"
Remember
Data itself is neutral
People choose how to present it
Even honest people make mistakes
Skepticism is a skill, not rudeness
The Full Data Cycle

Everything You've Learned

👁️
Sessions 1–2
Notice & Ask
📋
Sessions 3–4
Collect & Organize
📊
Sessions 5–6
Visualize & Interpret
🕵️
Session 7
Evaluate & Question
Session 8 = you run the whole cycle yourself — your own data project!
Vocabulary Check

Today's Key Words

Misleading
Giving a false impression, even if the numbers are technically accurate
Truncated axis
A chart axis that doesn't start at zero, making differences look larger than they are
Small sample
Too few people asked — results can't be trusted for a larger group
Biased question
A question that pushes toward one answer (revisited from Session 2)
Critical reader
Someone who questions what they see instead of accepting it immediately
Reliable data
Data that is trustworthy — collected fairly, from enough people, with honest questions
Session 7 Wrap-Up
🎉

You're a Data Myth Buster!

Interpret → Evaluate → (next: YOUR project!)
What you did today
Learned 3 ways data can mislead
Busted 4 misleading data myths
Learned the 3-question test for any chart
Became a critical, questioning data reader
Coming up in Session 8
Your own complete data project
Pick a question → collect → organize → visualize → present
The full data cycle from start to finish
Share your work with the class!
Data Science for Young Minds · Grade 3 · sdabagh.github.io/learn/data-science