Math / calculator
Sample Size Calculator
Plan a proportion survey sample using confidence, percentage-point error and an expected proportion, with finite population correction.
Calculator
Calculator
Estimate a proportion sample size.
Normal-approximation planning for independent random sampling. Enter percentage-point margin and expected percentage, with optional finite population correction.
Your result
Enter details to see the result.
Complete the inputs and calculate to see the main result and its supporting breakdown.
Example
Worked example
Example inputs
- Confidence level
- 95%
- Margin of error (percentage points)
- 5
- Expected population proportion (%)
- 50
Calculation breakdown
n0=z²p(1−p)/e²; n=n0/[1+(n0−1)/N] when N is supplied; required responses=ceil(n).
- Uncorrected continuous estimate
- 384.14588207
Example result
Required completed responses
385
Always rounds upward. No design effect or nonresponse allowance; proportions near 0 or 100% can need exact methods.
How to interpret the result
Required responses round the corrected continuous estimate upward, with at least one response. Add separate nonresponse and design-effect allowances when your study calls for them.
How to use this calculator
- Choose confidence and enter the permitted percentage-point margin.
- Supply an expected proportion, or use 50% for the conservative normal estimate.
- Optionally enter a finite population, then use the upward-rounded response count.
How it works
Formula
n0=z²p(1−p)/e²; n=n0/[1+(n0−1)/N] when N is supplied; required responses=ceil(n).
Why start with 50%?
p(1−p) is largest at 0.5, giving the largest normal-approximation sample for a fixed confidence and margin.
What this calculator does
Plan a proportion survey around a chosen confidence level, error margin and expected proportion before collecting responses.
Inputs you will need
Required or selected inputs
- Confidence level
- Margin of error (percentage points)
- Expected population proportion (%)
Optional inputs
- Population size (optional)
Assumptions
- Simple independent random sampling and a normal approximation are assumed; finite correction assumes sampling without replacement.
Limitations
- This plans completed responses, not invitations. Proportions must be strictly between 0% and 100%; clusters, bias and rare-event exact methods are excluded.
Useful next steps
For a mean interval after collecting a sample, use the different model in Confidence Interval Calculator.
References
NIST: confidence limits, normal and Student t distributions — Mean intervals use population SD with Z, or sample SD with T and n−1 degrees of freedom. P-values require an explicitly chosen distribution and tail. T evaluation uses the DLMF incomplete-beta continued fraction; survey planning uses the normal proportion approximation.
See how HandyChest handles units, validation, rounding, and testing, or Report a problem with this calculator.
Good to know
Frequently asked questions
How should I read the sample size result?
Required responses round the corrected continuous estimate upward, with at least one response. Add separate nonresponse and design-effect allowances when your study calls for them.
Keep going