1 / 15
Session 2 of 8
Research Questions
& Hypotheses
Today we learn how to turn curiosity into something testable — the foundation of all scientific thinking.
🔍 Data Science for Young Minds · Grade 5 — Data Detective
2 / 15
Today's Plan
What We're Doing Today
- 🎯 Hook — three questions that can't be answered with data (yet)
- ❓ What makes a question a research question?
- 🔬 Building a testable hypothesis
- 🃏 Hypothesis Formula Card — your new best tool
- ✂️ Sharpening Activity — fix 5 vague questions
- 💭 Null hypothesis — the skeptic's starting point
3 / 15
Opening Hook
Can You Answer This With Data?
Look at these three questions. Could you collect data to answer them?
- ❌ "Why do people like music?"
- ❌ "Are kids healthy?"
- ❌ "Is homework useful?"
These are interesting questions — but they're too vague, too broad, or too opinion-based to test with data. A Data Detective needs a sharper question.
4 / 15
Lesson 1
What Makes a Good Research Question?
❌ Too Vague / Not Testable
- Has no specific group or variable
- Can't be measured or counted
- Asks for an opinion, not a fact
- "Is exercise good?" — good for whom? how measured?
✅ Specific & Testable
- Names a specific group
- Has a measurable outcome
- Can be confirmed or denied by data
- "Do 5th graders who exercise 30+ min/day score higher on weekly math quizzes?"
A good research question is like a lock — and a hypothesis is the key that fits it.
5 / 15
Lesson 1
Variables — What Can Change?
A variable is anything that can differ between subjects or change over time. Good research questions always involve at least two variables.
Independent Variable
The thing you think causes the change (e.g., exercise amount)
Dependent Variable
The outcome you're measuring (e.g., quiz score)
Example Question
"Does sleep length affect memory test performance?"
Variables Here
Independent: sleep length. Dependent: memory test score.
When you see a research question, ask yourself: What is being changed? What is being measured?
6 / 15
Lesson 2
Building a Hypothesis
A hypothesis is an educated prediction about what you expect to find — before you look at the data.
- The "because" connects your prediction to prior knowledge or reasoning
- The "I predict" states a specific, measurable expected outcome
- A hypothesis can be wrong — and that's scientifically valuable
7 / 15
Lesson 2
From Vague to Testable
8 / 15
Activity Time!
Sharpen 5 Vague Questions
With your partner, turn each vague question into a testable hypothesis using your Formula Card.
- 1. "Do kids sleep enough?"
- 2. "Does music help you study?"
- 3. "Are phones bad for kids?"
- 4. "Do people eat healthier now?"
- 5. "Is exercise good for grades?"
⏱ You have 18 minutes. For each: name the variables, write the hypothesis, state your prediction.
9 / 15
🧠
Brain Break — Testable or Not?
Stand up for TESTABLE, stay seated for NOT TESTABLE.
"Do taller students jump higher?" · "Is kindness good?" · "Do students who eat breakfast focus better?" · "What is the best music?" · "Do cats sleep more than dogs?"
Notice: testable questions have specific, measurable answers!
10 / 15
Activity Debrief
Share Out — What Did You Write?
"Read your best hypothesis. What are the two variables? What's your predicted outcome? Could someone else collect data to test it?"
- Did your hypothesis have a "because"? That's what separates a guess from a prediction.
- Is your prediction specific enough? "Better grades" is vague. "5+ points higher average" is testable.
- Could two different people design the same study from your hypothesis? If yes — excellent!
11 / 15
Lesson 3
The Null Hypothesis — The Skeptic's Start
In science, the default assumption is: "Nothing is happening. There is no difference." This is the null hypothesis.
Null Hypothesis
- "Exercise has no effect on quiz scores"
- "Music makes no difference to study speed"
- "Sleep length doesn't affect memory test results"
Your Hypothesis (Alternative)
- "Exercise improves quiz scores"
- "Music makes students study faster"
- "More sleep = better memory scores"
Think of it like a court case: the null hypothesis is "innocent until proven guilty." Your job is to collect data that proves something IS happening.
12 / 15
Lesson 4
Why Predict BEFORE Collecting?
- 🧪 If you look at data first, your brain finds patterns it wants to see
- 🔒 Writing your prediction first "locks in" your honest expectation
- 🔬 Real scientists pre-register their hypotheses before collecting data
- 🎯 It makes your conclusion more trustworthy — you can't be accused of cherry-picking
In Session 4, we'll see what happens when scientists (or journalists!) don't do this. The results can be very misleading. Your hypothesis protects you from fooling yourself.
13 / 15
Write Time
Your Own Hypothesis
"Think of ONE topic you're curious about.
Write a full hypothesis using your Formula Card.
Name both variables. State your predicted outcome."
✍️ 6 minutes. Use your worksheet — Part 4. Choose something you actually care about. This hypothesis could become your Session 8 capstone project!
14 / 15
Vocabulary Check
Session 2 Key Terms
Hypothesis
A testable, educated prediction made before collecting data
Variable
Something that can change or differ between subjects
Testable
Can be confirmed or denied by collecting real data
Prediction
A specific, measurable expected outcome
Null Hypothesis
The default: "I expect no difference or relationship"
Expected Result
What you think the data will show — stated in advance
15 / 15
Session Close
The Detective's Hypothesis Rule
"A hypothesis isn't a guess — it's a commitment to honest testing."
- ✅ You can write a testable research question
- ✅ You can build a hypothesis using the Formula Card
- ✅ You understand what the null hypothesis means
- ✅ You know why you predict before you collect
Next session: Sampling — how do we collect data from a big group without surveying everyone?