Sample Size Calculator

Work out how many people you need before you collect data. For surveys and polls, enter the confidence level, margin of error and expected proportion, plus the population size if it is small. To estimate a mean, enter the expected standard deviation and the precision you need. For experiments that compare groups, the power-analysis tabs cover t-tests, ANOVA, two proportions, correlation and chi-square.

Survey sample size formula

n₀ = z² · p(1 − p) ÷ E² · with a finite population N: n = n₀ ÷ (1 + (n₀ − 1) ÷ N)

Here z is the critical value for your confidence level (1.96 for 95%), p is the expected proportion and E is the margin of error as a decimal. If you have no idea what p will be, use 50%. It gives the largest, safest sample.

Common survey sample sizes (95% confidence, p = 50%)

Margin of errorLarge populationPopulation 1,000Population 500
±10%978881
±5%385278218
±3%1,068517341
±1%9,604906476

Plan for non-response

The calculator gives the number of completed responses you need. Divide by your expected response rate to find how many people to invite: 385 responses at a 40% response rate means inviting 963 people. The table under the result does this for common rates.

What about Slovin’s and Yamane’s formulas?

Slovin’s formula n = N ÷ (1 + N·e²) is popular in some courses. It gives the same answer as the formula above only for 95% confidence with p = 50%, because it silently fixes both. The general formula lets you change them, which makes it the better choice.

Frequently asked questions

How many participants do I need for an experiment?

That depends on the effect size you want to detect and the power you need. Use the t-test, ANOVA or two-proportion tabs above, or the dedicated power analysis calculator.

Does a bigger population need a much bigger sample?

Surprisingly, no. Beyond about 20,000 people the required sample barely changes. A national poll of 1,000 people is as precise as a city poll of 1,000.