Sample Size Calculator
Estimate completed responses for a proportion from confidence, margin, prevalence and population inputs.
What would you like to do next?
Plan a proportion estimate before collecting responses
Select a confidence level, enter the desired margin of error and supply an expected proportion. The calculator reports the completed-response count from a normal-approximation formula and can apply a finite-population correction when a positive population size is entered. Using 50 percent gives the most conservative variance within this simple model, while a defensible prior estimate can produce a different requirement.
Add the design realities that the formula cannot see
The result assumes a simple random sample, one primary proportion and usable responses from everyone counted. Cluster sampling, weights, subgroup reporting, multiple outcomes, low response rates and expected exclusions can increase the invitations or completed cases needed. Treat the output as an initial planning quantity, document every adjustment and seek design-specific statistical review for consequential research.
Frequently asked questions
Why is 50 percent often used?
It produces the largest variance for a proportion and therefore a conservative initial sample size.
What does population size change?
A known finite population can reduce the required count through the finite-population correction.
Does the result include nonresponse?
No. Increase invitations separately for expected nonresponse or unusable records.