Free tool

Sample size calculator

Get the right sample size for surveys, qualitative interviews, and usability tests — with a plain-English explanation of the reasoning behind each recommendation.

Research type

Select the type of research you are planning.

Confidence level
Margin of error

Leave blank if your population is large or unknown.

The percentage of people you invite who will complete the survey.

Recommended sample size

385

completed responses

n = Z² × p(1−p) / e²

Why this number?

To be 95% confident that your results are accurate to within ±5%, you need 385 completed responses. This assumes your population is large or unknown, and uses the most conservative estimate of variability (p=0.5).

Important caveats

  • · This assumes simple random sampling. Stratified or cluster sampling will change your requirements.
  • · A 50% variability assumption (p=0.5) is used — the most conservative estimate. If you expect a higher or lower proportion, your required sample may be smaller.
  • · Non-response bias is not accounted for by sample size alone. A higher response rate from a smaller invitation list is generally better than a low response rate from a large one.