A familiar situation
Plan the approximate response count for a poll based on confidence and margin-of-error targets.
How to use this calculator
- Enter confidence level, margin of error, expected proportion and the remaining inputs.
- Select Calculate to see the answer and its breakdown.
- Keep a scenario, change an input and compare the results.
Use consistent units and check the assumptions below before applying the result.
How sample size is calculated
The large-population estimate is n₀=z²p(1−p)/e². For finite population N, apply n=n₀/[1+(n₀−1)/N]. The method assumes a simple random sample and a proportion outcome. At 95% confidence, 5% margin, and p=0.5, a large population needs about 385 responses.
n₀=z²p(1−p)/e²; finite population n=n₀/[1+(n₀−1)/N].At 95% confidence, 5% margin, and p=0.5, a large population needs about 385 responses.
Assumptions & useful context
The result depends on the values you enter and the method described above. Displayed numbers may be rounded; calculations use the unrounded values. Check that this model matches your situation before relying on it.
Frequently asked questions
Why is 50% the default proportion?
It maximizes p(1−p), producing a conservative largest sample estimate when the true share is unknown.
Does this guarantee a representative sample?
No; nonresponse and sampling design can introduce bias that the formula does not remove.