Sample Size Calculator

Your details

Find the sample you need, or check what margin a size you already have gives you.
Proportion for survey questions with a percentage answer; mean for numeric measurements like height or spend.
How often the true value falls inside your interval across many repeated studies. 95% is the standard default.
The +/- accuracy you can tolerate, in percentage points for proportions or in the same unit as your measurement for means.
%
50% is the safe conservative default. It maximises the required sample, so you are never under-powered.
%
How far off from the true mean you can tolerate, in the same unit as the measurement.
Total size of the group you are studying. Leave at 0 for an unknown or very large population.
Required sample size
385responses
Without population correction385
385 responses
Very small<100Adequate100-385Robust385-1000Very robust1000+

You need about 385 responses.

  • At 95% confidence, aim for 385 completed responses.
  • A 50% expected proportion is the conservative choice. It maximises the sample, so a different true split only makes your survey more precise.
  • Halving the margin of error roughly quadruples the sample you need, so set the loosest margin your decision can tolerate.

Next stepAccount for non-response: toggle "Adjust for non-response" and enter your expected completion rate to see how many invitations to send.

 = Powered by OnlyCalculators