Formulas
n₀ = z² × p(1 − p) ÷ e²
With a known population N: n = n₀ ÷ (1 + (n₀ − 1) ÷ N)
Margin of error = z × √(p(1 − p) ÷ n)
z is 1.645 for 90%, 1.96 for 95% and 2.576 for 99% confidence. Using p = 50% gives the largest (safest) sample.
Sample sizes for large populations (p = 50%)
| Margin of error | 95% confidence | 99% confidence |
|---|---|---|
| ±1% | 9604 | 16588 |
| ±2% | 2401 | 4147 |
| ±3% | 1068 | 1844 |
| ±4% | 601 | 1037 |
| ±5% | 385 | 664 |
| ±10% | 97 | 166 |
Example
For 95% confidence and a ±5% margin, you need 385 responses from a large population. If the whole population is only 1,000 people, 278 responses are enough.
Tips
- Plan for non-response: if you expect 20% of people to answer, invite five times as many.
- A bigger sample doesn't fix a biased one; who you ask matters as much as how many.
- For a confidence interval around an average, use the confidence interval calculator.
Frequently asked questions
Why is 385 such a common sample size?
It's what you need for 95% confidence and a ±5% margin of error in a large population.
Does population size matter?
Only when the sample is a sizeable share of the population; for large populations it barely changes the answer.
What proportion should I use?
Use 50% if you have no estimate; it gives the most conservative sample size.
Written and checked by Vijay. Last updated .