Saved

Press the star next to a tool's name to save it.

All my tools

Uncertainty Calculator

Use this uncertainty calculator for absolute, relative and percent uncertainty, and to propagate errors through sums, differences, products, quotients and powers.

Restored your last values ·

Enter a measured value and its absolute uncertainty to get the relative and percent uncertainty, or the other way round. For error propagation, enter two values with their uncertainties and choose sum, difference, product, quotient or a power.

Everything is worked out in your browser. Nothing you type is sent anywhere.

–

–

Choose what to work out, fill in the boxes, then press Calculate.

See full result

An uncertainty calculator turns a measurement and its uncertainty into a relative or percent uncertainty, and works out the uncertainty of a result calculated from several measurements. It covers the absolute uncertainty, and the error propagation of sums, differences, products, quotients and powers.

How to use the uncertainty calculator

  1. Choose what to work out. For a single measurement, enter the value and its absolute or percent uncertainty.
  2. For error propagation, choose the operation and enter A and B with their uncertainties.
  3. Press Calculate.

Absolute and relative uncertainty

The Wikipedia article on propagation of uncertainty says that uncertainty may be defined by the absolute error Δx, or by the relative error Δx ÷ x, which is usually written as a percentage.[1] It adds that most commonly the uncertainty on a quantity is expressed as a standard deviation σ, and that the value and its error are then written as an interval x ± u.[1] Enter 50 with an absolute uncertainty of 2 and the calculator shows 4 percent. Enter 50 with 4 percent and it gives 50 ± 2. We use the absolute value of x for the relative form.

Sums and differences

For f = A + B or f = A − B with uncorrelated variables, the page gives σf squared = σA squared + σB squared.[1] The uncertainties add in quadrature, which means that you square them, add them and take the square root. For A = 10 ± 0.3 and B = 5 ± 0.4, the sum is 15 and the uncertainty is the square root of 0.09 + 0.16, which is 0.5. The difference is 5 ± 0.5.

Products and quotients

For a product f = AB, the page gives the relative uncertainty of the result as (σf ÷ f) squared ≈ (σA ÷ A) squared + (σB ÷ B) squared for uncorrelated variables, and gives the same form for a quotient.[1] With A = 10 ± 0.3 and B = 5 ± 0.4, the relative uncertainties are 3 percent and 8 percent, and in quadrature they give 8.54 percent. The product is 50 ± 4.27, and the quotient is 2 ± 0.171. The page also shows the example of Ohm's law, R = V ÷ I, which gives σR = R × √((σV ÷ V)² + (σI ÷ I)²).[1]

Powers

For f = aA to the power b, the page gives σf ≈ |f × b × σA ÷ A|.[1] For A = 10 ± 0.3 and n = 2, f = 100 and the uncertainty is 100 × 2 × 0.03 = 6. A power multiplies the relative uncertainty by the exponent, so squaring doubles it, and taking a square root halves it.

The general rule

For any function the page says that the function is usually linearised with a first-order Taylor expansion, and gives the simplified form for independent variables as the square root of the sum of (∂f ÷ ∂x)² σx² over all variables.[1] The sums, products and powers above are special cases. This calculator covers those cases directly and does not do arbitrary functions.

Correlated errors and limits

The page says that if the uncertainties are correlated, the covariance must be taken into account, with σij = ρij σi σj.[1] It also says that error estimates for non-linear functions are biased because of the truncated expansion, and that the linear approximation must be close to the function over the range of the uncertainties.[1] If A and B are not independent, for example when A − A appears in a formula, the independent formulas give the wrong answer. Use them only for separate measurements.

Worked examples

A length of 12.5 cm measured with a ruler to ± 0.1 cm has a relative uncertainty of 0.8 percent. A rectangle of 12.5 ± 0.1 by 4.0 ± 0.1 cm has an area of 50 cm² with relative uncertainties of 0.8 percent and 2.5 percent, which combine in quadrature to about 2.62 percent, or ± 1.31 cm². These are our own examples worked from the formulas above.

Choosing the right rule

Look at the operation that produced your result. If the result comes from adding or subtracting measurements, work with absolute uncertainties. If it comes from multiplying or dividing, work with relative uncertainties. If a quantity is raised to a power, multiply its relative uncertainty by the exponent. For a longer chain of operations, break the calculation into steps, find the uncertainty at each step and carry it forward. Each step uses one of the rules that the calculator provides.

Significant figures

State the uncertainty with one or two significant figures and round the result to the same decimal place. For 50 ± 4.272, write 50 ± 4. The pages we read give no rule for rounding, so this is common practice and our own advice. A result written with more digits than its uncertainty supports suggests a precision that the measurement does not have.

Repeated measurements

When you measure the same quantity several times, the uncertainty of the mean is the standard deviation of the readings divided by the square root of their number. The page states this for the mean of n items.{1} So ten readings give an uncertainty about a third of one reading, a gain that grows slowly as you take more readings.

Common mistakes

  • Adding uncertainties directly. Independent errors add in quadrature, not linearly.
  • Adding relative uncertainties for sums. Use absolute uncertainties for sums and differences.
  • Using the formulas for correlated errors. Include the covariance term.

What next

Use this often? Add Uncertainty to your Home screen.

Frequently asked questions

What is an uncertainty calculator?

A tool that finds absolute, relative and percent uncertainty and propagates the uncertainties of measurements through a calculation.

How do I calculate percent uncertainty?

Divide the absolute uncertainty by the measured value and multiply by 100. A value of 50 ± 2 has 4 percent uncertainty.

How do uncertainties combine in a sum?

In quadrature: the uncertainty of A + B is the square root of σA² + σB², for independent errors.

How do uncertainties combine in a product?

The relative uncertainties add in quadrature: (σf ÷ f)² ≈ (σA ÷ A)² + (σB ÷ B)².

What if the errors are correlated?

Then the covariance terms must be included. The simple formulas assume independent errors.

What went wrong?

We only receive the reason you pick, never what you typed into the tool.

Thanks for your feedback.

More Math Tools