Lesson 2: The Problem With Small Samples

About 30 minutes — Discussion-based lesson

What You Will Learn

This lesson covers:

Why 3 people is not enough to make conclusions about 300
How sample size affects reliability
The coin flip example: small samples give weird results
When is a sample big enough?

Why 3 people is not enough to make conclusions about 300

This section covers the key ideas about why 3 people is not enough to make conclusions about 300. Discuss with your group or family and explore the concepts together.

How sample size affects reliability

This section covers the key ideas about how sample size affects reliability. Discuss with your group or family and explore the concepts together.

The coin flip example: small samples give weird results

This section covers the key ideas about the coin flip example: small samples give weird results. Discuss with your group or family and explore the concepts together.

When is a sample big enough?

This section covers the key ideas about when is a sample big enough?. Discuss with your group or family and explore the concepts together.

Check Your Understanding

1. Why are small samples unreliable?

Answer: With few responses, one unusual answer can dramatically change results. If you ask 3 people and 1 loves broccoli, it seems like 33% love broccoli — but ask 100 and it might be 5%.

2. What is the coin flip example?

Answer: Flip a coin 5 times — you might get 4 heads and think the coin is unfair. Flip it 100 times and you will get close to 50-50. Small samples are noisy; large samples are reliable.

3. When is a sample big enough?

Answer: There is no magic number, but more is generally better. For a class survey, try to ask at least 20-30 people. For bigger questions, you need bigger samples.

4. Can a large sample still be wrong?

Answer: Yes, if it is biased! Asking 1,000 people who all live in one city does not tell you about the whole country. Size matters, but so does who you ask.

Key Takeaways

Why 3 people is not enough to make conclusions about 300
How sample size affects reliability
The coin flip example: small samples give weird results
When is a sample big enough?

Ready for More?

Next Lesson

Continue to Lesson 3: Biased Data

Start Lesson 3

Session Home

Review all lessons and materials.

Back to Session Home