Sign inGet started

Free tool · no sign-up

Survey sample size calculator

How many responses do you need? Pick a confidence level and margin of error, add your population if you know it, and allow for the share of people who will actually complete.

Your survey

Confidence level

5% is the usual choice; 3% for close comparisons.

For example the number of customers, employees or voters in a ward.

The share of invited people who complete. Use 100% if you only count completes.

Advanced: expected proportion

Leave at 50% unless earlier research tells you otherwise; 50% gives the safest (largest) sample.

Completed responses needed

385

at 95% confidence, ±5% margin of error

Already have responses?

Margin of error: ±6.9% at 95% confidence.

Collect them in 12 Indian languages

Set your survey's target to this number and let respondents answer in Hindi, Tamil, Bengali and 9 more languages.

See how SurveySetu works

How the calculation works

The calculator uses Cochran's formula for estimating a proportion from a simple random sample, with the finite population correction:

  • n₀ = z² × p(1 − p) ÷ e² — completes for a very large population, where z is 1.96 at 95% confidence, p the expected proportion and e the margin of error.
  • n = n₀ ÷ (1 + (n₀ − 1) ÷ N) — the correction for a population of N people.
  • Invitations = n ÷ response rate — how many people to invite so that n of them complete.

It assumes respondents are chosen at random from the population. Samples from a convenience source (for example a WhatsApp group) can be biased whatever their size, so describe who answered when you report results.

Sample size questions

What sample size do I need for a survey?

For a large population, 385 completes give a ±5% margin of error at 95% confidence, and about 1,068 give ±3%. Smaller populations need fewer. Enter your own confidence level, margin of error and population above to get your number.

What confidence level and margin of error should I use?

95% confidence with a ±5% margin of error is the usual default for opinion and market research. Use ±3% when small differences matter, or 90% confidence when a rough read is enough and responses are expensive.

What if I don't know the population size?

Leave it empty. Once a population is more than about 100,000 people, its exact size barely changes the sample you need, so the calculator treats it as unlimited.

Why does the response rate matter?

The sample size is the number of completed responses you need. If only a share of the people you invite will complete, you need to invite more: at a 20% response rate, 385 completes means inviting about 1,925 people.

Does this work for subgroups?

Each subgroup you want to report on separately (for example each state, or men and women) needs enough completes on its own. Calculate the sample for one subgroup and multiply by the number of subgroups.

Is a bigger sample always better?

Only if it is the right people. A large sample from the wrong audience, or with careless answers, is still biased. Targeting, screening and quality checks matter as much as the number of completes.

Keep exploring