Overview

Sampling error is the difference you can expect between a sample's result and the true population value, simply because you measured only part of the group. This gives the standard error and the margin of error.

Enter the sample's standard deviation and size, and a confidence code: 1 for 90%, 2 for 95% or 3 for 99%. If the population is small, enter its size. Otherwise enter 0.

Worked examples

InputsResult
Sample Standard Deviation12Sample Size (n)64Confidence level95%Population Size (0 if large)0Standard error: 1.5Finite population correction: 1Margin of error: 2.94
Show 2 more examples
Sample Standard Deviation12Sample Size (n)64Confidence level95%Population Size (0 if large)500Standard error: 1.4Finite population correction: 0.93Margin of error: 2.75
Sample Standard Deviation0.5Sample Size (n)1,000Confidence level95%Population Size (0 if large)0Standard error: 0.02Finite population correction: 1Margin of error: 0.03

Standard error by sample standard deviation

With the calculator's starting values, the standard error is 1.5. The table shows how it changes with the sample standard deviation.

Sample Standard DeviationStandard error
60.75
81
101.25
12starting value1.5
141.75
162
182.25

Other inputs: Sample Size: 64 n · Population Size: 0 0 if large. 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. If your sample is a meaningful share of a known population, the finite population correction shrinks it. The margin of error is the standard error times a z value of 1.645, 1.96 or 2.576.

The formula

Margin of error = z × s ÷ √n × √((N − n) ÷ (N − 1))
sSample standard deviation
nSample size
NPopulation size, if known
zValue for your chosen confidence level

Practical tips

  • With s = 12 and n = 64, the standard error is 1.5, and the 95% margin of error is 2.94.
  • To halve the margin of error, you need four times the sample, because n sits under a square root.
  • This only covers random sampling error. It doesn't account for biased samples or poor questions.

Frequently asked questions

When should I enter a population size?

When your sample is more than about 5% of the population. For large populations, the correction barely matters.

What's the difference between standard error and margin of error?

The standard error is the typical sampling variation. The margin of error scales it to a chosen confidence level.

Is this the same formula for poll percentages?

The idea is the same, but for a percentage p, the standard deviation is √(p(1 − p)).