Free research tool
Margin of Error Calculator
Find out how precise a survey result is. Enter your sample size and result to get the ± margin and the range the true value most likely sits in. 1,000 responses give ±3.1% at 95% confidence.
Free, no sign-up. Last reviewed against source documents on .

Number of completed responses.
e.g. 62% said yes. Use 50% for the worst case across all questions.
Only matters if you surveyed a big share of a small group.
Raise above 1.0 if the data is weighted.
Margin of error table
Margin of error for a 50/50 result (the widest case) by sample size, for a large population.
| Sample size | 90% confidence | 95% confidence | 99% confidence |
|---|---|---|---|
| 50 | ±11.6% | ±13.9% | ±18.2% |
| 100 | ±8.2% | ±9.8% | ±12.9% |
| 200 | ±5.8% | ±6.9% | ±9.1% |
| 300 | ±4.7% | ±5.7% | ±7.4% |
| 400 | ±4.1% | ±4.9% | ±6.4% |
| 500 | ±3.7% | ±4.4% | ±5.8% |
| 750 | ±3.0% | ±3.6% | ±4.7% |
| 1,000 | ±2.6% | ±3.1% | ±4.1% |
| 1,500 | ±2.1% | ±2.5% | ±3.3% |
| 2,000 | ±1.8% | ±2.2% | ±2.9% |
| 3,000 | ±1.5% | ±1.8% | ±2.4% |
| 5,000 | ±1.2% | ±1.4% | ±1.8% |
Quadrupling the sample only halves the margin: 250 → 1,000 takes ±6.2% to ±3.1%.
How the calculation works
Worked example: 62% of 400 respondents said yes. At 95%: 1.96 × √(0.62 × 0.38 ÷ 400) = 0.048, so 62% ±4.8%, a range of 57.2% to 66.8%.
Comparing two results rather than reading one? The margins of two groups don’t simply add up, so use the significance calculator. Planning a new study? Start with the sample size calculator.
Questions, answered
What is the margin of error in a survey?
The margin of error is how far a survey result could be from the true population value because only a sample was asked. A result of 40% with a ±3% margin at 95% confidence means that if you repeated the survey many times, about 95 in 100 such intervals (37% to 43%) would contain the true value.
What is the margin of error for 1,000 respondents?
±3.1% at 95% confidence for a 50/50 result. For a result of 20% or 80% it narrows to ±2.5%. At 99% confidence it is ±4.1%.
How do you calculate margin of error?
Margin of error = z × √(p × (1 − p) ÷ n), where z is 1.96 for 95% confidence, p is the result as a decimal and n is the sample size. Multiply by √DEFF for weighted data, and by √((N − n) ÷ (N − 1)) when you sampled a big share of a small population N.
Does the margin of error apply to subgroups?
No. Each subgroup has its own, larger margin. In a survey of 1,000 (±3.1%), a subgroup of 200 women aged 18 to 24 has a margin of about ±6.9%. Report subgroup results with their own margin, or flag small bases.
Is a 5% margin of error good?
±5% is fine for most tracking and directional decisions. Headline figures in published polls usually aim for ±3% or better. If two options are separated by less than the margin, do not call a winner without a significance test.
Does margin of error cover all survey error?
No. It only covers random sampling error. It says nothing about a biased sample, leading questions, non-response or fraudulent respondents, which can be much larger. For opt-in online panels, which are not probability samples, AAPOR says a "credibility interval" may be the appropriate measure, and urges caution in interpreting it.
Sources and method
Normal approximation for a proportion, with optional finite population correction and design effect. Intervals are clipped to 0 to 100%. All figures on this page come from the same tested code as the calculator.
More free research tools

Sample Size Calculator
Completes you need, and how many people to reach after incidence

Significance Calculator
Is the gap between two groups real? Percentages or averages

NCCS Calculator
Classify any Indian household into NCCS A1 to E3 in two questions

ISEC Calculator
MRSI’s new socio-economic classes, ISEC 1 to 12, for any household
Need tighter numbers? Get more completes.
Build the survey in PollPe and reach respondents across India by WhatsApp, QR or link.