Overview

The margin of error tells you how far a sample's average might be from the true population average. It's the '±' you see in reported survey results.

Enter the sample's standard deviation and size, then a confidence code: 1 for 90%, 2 for 95% or 3 for 99%.

Worked examples

InputsResult
Sample Standard Deviation12.4Sample Size (n)120Confidence level95%Margin of error: 2.22Confidence level: 95%
Show 2 more examples
Sample Standard Deviation0.5Sample Size (n)1,000Confidence level95%Margin of error: 0.03Confidence level: 95%
Sample Standard Deviation15Sample Size (n)40Confidence level99%Margin of error: 6.11Confidence level: 99%

Margin of error by sample standard deviation

With the calculator's starting values, the margin of error is 2.22. The table shows how it changes with the sample standard deviation.

Sample Standard DeviationMargin of error
81.43
101.79
122.15
12.4starting value2.22
142.5
162.86
183.22

Other inputs: Sample Size: 120 n. Change them in the calculator above for your own numbers.

How it works

The standard error is the standard deviation divided by the square root of the sample size. Multiply it by a z value of 1.645, 1.96 or 2.576 for your confidence level. It uses z values throughout, so it's a close approximation for smaller samples, where a t value would be slightly larger.

The formula

MOE = z × s ÷ √n
z1.645 for 90%, 1.96 for 95%, 2.576 for 99%
sSample standard deviation
nSample size

Practical tips

  • With s = 12.4 and n = 120, the 95% margin of error is about 2.22.
  • For a yes/no poll at 50%, 1,000 responses gives about ±3.1%. Enter s = 0.5 and n = 1000, then multiply the result by 100.
  • It measures random error only. It says nothing about biased questions or who chose to reply.

Frequently asked questions

How do I reduce the margin of error?

Increase the sample size. Quadrupling it halves the margin.

What does 95% confidence mean?

If you repeated the survey many times, about 95% of the intervals built this way would contain the true value.

Do I need a larger sample for a large population?

Barely. Once a population is large, the sample size matters far more than the population size.