Skip to content
Omnikalc

Tolerance interval calculator (normal data)

Enter the sample mean, the sample standard deviation and the sample size, then choose the share of the population to cover and the confidence, to get the tolerance factor and the lower and upper tolerance limits.

Updated on
Your inputs

Value out of range

Value out of range

Value out of range

Results

Two-sided tolerance factor k

2.752

Moderate inflation compared with the naive interval
Lower tolerance limit
38.991
Upper tolerance limit
61.009
Half-width of the interval
11.009
Inflation factor against the naive interval
1.404 ×
One-sided tolerance factor (approximation)
2.378
Coverage chosen
95 %
Confidence chosen
95 %

The tolerance interval calculator gives the limits that contain a chosen share of a normal population with a chosen confidence, from the sample mean, standard deviation and size.

How the calculation works

  • Enter the mean, the standard deviation and the sample size.
  • Choose the share to cover and the confidence.
  • The calculator gives the tolerance factor, the lower and upper limits, the half-width and the inflation against the naive interval.

Example

Twenty parts have a mean of 50 and a standard deviation of 4. For 95 % coverage with 95 % confidence, k = √(19 × 1.05 × 1.96² / 10.12) = 2.752. The limits are 50 ± 2.752 × 4, so from 38.991 to 61.009. The interval is 1.404 times wider than the naive mean ± 1.96 standard deviations, which ignores the uncertainty on the estimates.

Why it matters

The calculator helps quality engineers who set specification limits from a sample, laboratory analysts who define reference ranges, and students who learn the difference between confidence, prediction and tolerance intervals.

Good to know

The method assumes that the data come from a normal distribution, so check normality first with a plot or a test. With a small sample, the interval is much wider than the naive one, and only more data will shrink it. For non-normal data, transform them first or use a distribution-free interval, which needs a larger sample. The one-sided factor applies when only an upper or a lower limit matters, for example a maximum impurity level, and is computed with an approximation that is close to exact tables for samples above 10. Tolerance intervals are common in pharmaceutical, aerospace and reliability standards.

The confidence interval calculator brackets means, the normal distribution calculator gives probabilities, the process capability calculator compares spread with specifications and the margin of error calculator sizes surveys.

Frequently asked questions

What is a tolerance interval?

It is an interval that, with a stated confidence, contains at least a given share of the whole population, not just the mean. A 95 %/95 % interval covers at least 95 % of the values with 95 % confidence.

How does it differ from a confidence interval?

A confidence interval brackets a parameter, such as the mean. A prediction interval brackets one new value. A tolerance interval brackets a share of all values, so it is wider. The factor k grows when the sample is small.

How is k computed here?

The calculator uses Howe’s approximation: k = √((n − 1)(1 + 1/n) z² / χ²), where z is the normal quantile for the central share and χ² the lower chi-square quantile with n − 1 degrees of freedom at one minus the confidence. It matches published tables to three digits.

Was this calculator helpful?

Thanks for your feedback!

Results are estimates provided for information purposes. They do not replace professional advice.