Margin of Error

A common survey precision metric. Assumes a simple random sample and large-sample normal approximation.

MoE in percentage points = z × sqrt(p(1−p)/n) × 100, with p as a proportion.

Tip: Keep “Sample Proportion (%)” and “Sample Size (n)” on the same basis (period, units, and population) before calculating Margin of Error.

Cluster: Statistics hub · Percentage error · Percentage guide

Margin of error (for a proportion) estimates half-width of a Wald-style interval in percentage points.

Enter sample proportion (%), sample size n, and z critical value (e.g. 1.96 for ~95%).

%
Estimated proportion as percent
Number of respondents
e.g. 1.96 for ~95%

Margin of Error

Understanding Margin of Error

How we calculate. MoE in percentage points = z × sqrt(p(1−p)/n) × 100, with p as a proportion. The form uses the same arithmetic as the worked examples on this page. See our methodology and accuracy policy.

Real-world scenario: A typical Margin of Error case uses sample proportion (%) 50 and sample size (n) 400. Enter the same figures below to reproduce the worked path.

What is Margin of Error?

A common survey precision metric. Assumes a simple random sample and large-sample normal approximation.

  • p as percent in the form (converted internally)
  • z ≈ 1.96 for ~95% confidence
  • Output is percentage points

The Formula

Margin of Error (Proportion)
MoE (pp) = z × √(p(1−p)/n) × 100

Worked Example

Scenario: p = 50%, n = 400, z = 1.96.
Step 1: p = 0.5
Step 2: √(0.25/400) = 0.025
Step 3: MoE = 1.96 × 0.025 × 100 ≈ 4.90 pp
Answer: Margin of error is about 4.90 percentage points.

Common Use Cases

  • Poll reporting: ±MoE
  • Sample size planning: precision targets
  • Survey design: n vs width tradeoffs

Pro Tips

  • Use p=50% for conservative MoE
  • Finite populations may need corrections
  • Design effects can widen MoE

Limitations: Margin of Error results are educational statistics aids—not formal statistical certification. Confirm assumptions (SRS, large-sample approximations, labeling) for your analysis.

FAQ

Is MoE the same as confidence level?

No. Confidence level chooses z; MoE is the resulting half-width in percentage points.

What if n is 0?

MoE is undefined—enter a positive sample size.

Authoritative References

For statistical definitions and survey practice, consult:

  • NIST ITL — measurement and statistical guidance
  • AAPOR — survey response-rate standards
  • NCHS — public health statistics context