Sample Size Calculator
Run the Sample Size Calculator MicroSim Fullscreen
About This MicroSim
"Acorn for your thoughts?" Sylvia muses. "When you're planning a study, one of the most important questions is: How many individuals do I need to sample? Too few and your interval is too wide to be useful. Too many and you've wasted resources. Let's find that sweet spot!"
This calculator helps determine the required sample size to achieve a desired margin of error for a confidence interval about a proportion.
The Formula
Starting with the margin of error formula and solving for n:
The p Problem
We need p-hat to calculate n, but we don't have data yet! Two solutions:
- Conservative approach: Use p = 0.5 (maximizes sample size, guarantees the ME)
- Prior estimate: Use a value from previous research or pilot study
How to Use
- Set desired margin of error using the slider (1% to 10%)
- Choose confidence level (90%, 95%, or 99%)
- Select proportion approach:
- Conservative (p = 0.5) for guaranteed results
- Custom to use a prior estimate
- Read the required sample size (always rounded UP)
Key Insights
"My tail's tingling - we're onto something!" Sylvia observes:
- Halving the margin of error requires quadrupling the sample size! (because ME is under a square root)
- p = 0.5 is the safest choice when you don't know the true proportion
- Higher confidence requires larger samples (larger z* value)
- Always round UP to ensure you meet your ME goal
Sample Size Rules of Thumb
| Desired ME | n (95% conf, p=0.5) | Example Use |
|---|---|---|
| 10% | 97 | Quick informal poll |
| 5% | 385 | Standard survey |
| 3% | 1,068 | News organization poll |
| 2% | 2,401 | High-precision research |
| 1% | 9,604 | Very precise study |
Lesson Plan
Learning Objectives
By the end of this activity, students will be able to:
- Calculate the sample size needed for a given margin of error
- Explain why p = 0.5 is the conservative choice
- Understand the quadrupling rule for halving margin of error
- Make informed decisions about study design trade-offs
Target Audience
- AP Statistics students (high school)
- Introductory statistics college students
- Anyone designing surveys or studies
Prerequisites
- Understanding of margin of error
- Concept of confidence intervals
- Basic algebra (solving equations)
Classroom Activities
Activity 1: The Quadrupling Discovery (10 minutes)
- Set confidence = 95%, conservative p = 0.5
- Record n for ME = 4%
- Record n for ME = 2%
- Verify: Is the second n about 4 times the first?
- Try ME = 1% - is it 4x the 2% sample size?
Activity 2: Poll Planning (15 minutes)
Scenario: Your school newspaper wants to poll students about a new policy.
- If you can only survey 100 students, what ME can you achieve?
- If you need ME = 5%, how many students must you survey?
- If you want ME = 3% at 99% confidence, how many?
- Discuss: What are realistic constraints?
Activity 3: Conservative vs. Estimated (10 minutes)
- Use conservative (p = 0.5), ME = 3%, 95% confidence - note n
- Switch to custom p = 0.20 (prior research says about 20% will answer "yes")
- How much smaller is the required sample size?
- Discuss: When is it safe to use a prior estimate?
"Time to squirrel away this knowledge!" Sylvia concludes. "Planning sample size before collecting data is one of the most practical skills in statistics. It saves time, money, and headaches!"
Assessment Questions
-
A researcher wants a margin of error of 4% at 95% confidence with no prior estimate of p. What sample size is needed?
-
If a poll uses n = 1,000 at 95% confidence with p = 0.5, what is the expected margin of error?
-
Why does using p = 0.5 give a "conservative" (larger) sample size?
-
A marketing team has budget for 500 surveys. What margin of error can they expect at 95% confidence?
References
- Chapter 15: Confidence Intervals - Concepts: Determining Sample Size
- Wikipedia: Sample size determination