📖 What is Absolute Uncertainty?
Absolute uncertainty is a quantitative expression of the doubt in a measurement or calculated result. When a physicist reports a length as 5.23 ± 0.04 cm, the ±0.04 cm is the absolute uncertainty - it defines the range within which the true value is expected to lie. Every measurement contains some uncertainty, and reporting a number without an uncertainty is scientifically incomplete.
There are two sources of uncertainty in experimental science. Measurement uncertainty arises from the instrument's limitations: a digital balance reading to 0.1 g has an inherent uncertainty of ±0.05 g. Propagated uncertainty arises when measured quantities are combined mathematically - if you calculate area from length and width, the uncertainty in each measurement contributes to the uncertainty in the area.
The rules for propagating uncertainty depend on the mathematical operation. For addition and subtraction, absolute uncertainties combine in quadrature: Δz = √(Δx² + Δy²). This is more realistic than simply adding the uncertainties because it accounts for the statistical independence of the errors. For multiplication and division, relative uncertainties combine: Δz/z = √((Δx/x)² + (Δy/y)²). For a quantity raised to a power n, the relative uncertainty scales as n·|Δx/x|.
Understanding uncertainty propagation is essential in physics, chemistry, engineering, and any quantitative science. It tells you which measurement needs to be improved to reduce overall experimental error, and it prevents false precision - reporting more decimal places than the uncertainty justifies.
📐 Formulas
Single Measurement - Standard Deviation Method: Δx = σ / √n, where σ is the sample standard deviation and n is the number of measurements.
Single Measurement - Half-Range Method: Δx = (x_max − x_min) / 2
Addition or Subtraction (z = x ± y): Δz = √(Δx² + Δy²) - absolute uncertainties combine in quadrature.
Multiplication or Division (z = x·y or x/y): Δz/z = √((Δx/x)² + (Δy/y)²), so Δz = z · √((Δx/x)² + (Δy/y)²)
Power Rule (z = xⁿ): Δz/z = |n| · (Δx/x), so Δz = |n| · xⁿ⁻¹ · Δx = |n| · (z/x) · Δx
Relative Uncertainty: δx = Δx / |x| (dimensionless ratio, often expressed as %)
Result Notation: z ± Δz, rounded so that Δz has 1–2 significant figures and z matches the same decimal place.
All variables: x, y = measured values; Δx, Δy = absolute uncertainties; z = calculated result; n = exponent (power rule) or number of measurements (single); σ = sample standard deviation.
📖 How to Use This Calculator
1
Select the Propagation Mode that matches your situation: Single Measurement for one quantity with repeats, Addition/Subtraction for sums or differences, Multiplication/Division for products or ratios, or Power Rule for exponential expressions.
2
Enter the required values and their uncertainties. For single measurements, choose whether you have the standard deviation and count (σ/√n method) or only the min and max observed values (half-range method).
3
Click Calculate Uncertainty. The results show the absolute uncertainty Δz, the relative uncertainty as a percentage, the result value z, and the final result in standard ± notation.
💡 Example Calculations
Example 1 - Length measurement (single measurement, half-range)
1
A student measures a rod five times and gets: 24.7, 24.8, 24.7, 24.9, 24.8 cm. Min = 24.7, Max = 24.9, Mean = 24.78 cm.
2
Half-range: Δx = (24.9 − 24.7) / 2 = 0.1 cm. Result: 24.78 ± 0.10 cm (rounded to match uncertainty decimal place).
3
Relative uncertainty: 0.1 / 24.78 = 0.40% - very good precision for a simple ruler measurement.
Result = 24.78 ± 0.10 cm (relative uncertainty: 0.40%)
Try this example →Example 2 - Area calculation (multiplication propagation)
1
Length: x = 5.23 ± 0.04 cm. Width: y = 3.10 ± 0.02 cm. Area: z = x × y = 16.213 cm².
2
Relative uncertainties: Δx/x = 0.04/5.23 = 0.00765. Δy/y = 0.02/3.10 = 0.00645.
3
Propagated relative: √(0.00765² + 0.00645²) = 0.01001. Δz = 16.213 × 0.01001 = 0.16 cm². Result: 16.21 ± 0.16 cm².
Result = 16.21 ± 0.16 cm²
Try this example →Example 3 - Density measurement (division then power)
1
Mass m = 45.2 ± 0.1 g. Volume V = 18.0 ± 0.3 cm³. Density ρ = m/V = 2.511 g/cm³.
2
Relative uncertainties: Δm/m = 0.1/45.2 = 0.00221. ΔV/V = 0.3/18.0 = 0.01667.
3
Δρ/ρ = √(0.00221² + 0.01667²) = 0.01682. Δρ = 2.511 × 0.01682 = 0.042 g/cm³. Result: 2.51 ± 0.04 g/cm³. Volume uncertainty dominates - improving the volume measurement would reduce overall uncertainty most effectively.
Result = 2.51 ± 0.04 g/cm³
Try this example →Example 4 - Physics lab: kinetic energy (power rule)
1
Velocity v = 3.00 ± 0.05 m/s. Kinetic energy KE = ½mv² - velocity appears squared (n = 2).
2
Power rule: Δv/v = 0.05/3.00 = 1.667%. Propagated relative uncertainty = 2 × 1.667% = 3.333%.
3
With m = 2.00 kg: KE = 9.00 J. ΔKE = 9.00 × 0.03333 = 0.30 J. Result: 9.00 ± 0.30 J. The squared velocity doubles the relative contribution of any velocity uncertainty.
Result = 9.00 ± 0.30 J (relative uncertainty: 3.33%)
Try this example →❓ Frequently Asked Questions
What is the difference between absolute and relative uncertainty?+
Absolute uncertainty (Δx) is expressed in the same units as the measurement - for example, 5.23 ± 0.04 cm means the uncertainty is 0.04 cm. Relative uncertainty (also called fractional uncertainty) is the ratio Δx/x, often expressed as a percentage: 0.04/5.23 = 0.76%. Absolute uncertainty tells you the size of the error in real terms; relative uncertainty tells you how significant that error is compared to the measurement. A 0.04 cm uncertainty is negligible if x = 1000 cm but enormous if x = 0.05 cm.
What is the difference between measurement uncertainty and propagated uncertainty?+
Measurement uncertainty comes from the instrument or the act of measuring - for example, a ruler with 1 mm resolution has an absolute uncertainty of ±0.5 mm. Propagated uncertainty arises when you combine two or more measured quantities in a calculation. For example, if you measure length and width separately and then multiply to get area, the uncertainty in area must be calculated from the uncertainties in length and width using propagation rules. This calculator handles both types.
Why do we add uncertainties in quadrature (square root of sum of squares) rather than just adding them directly?+
Direct addition of uncertainties assumes both errors are at their maximum simultaneously and in the same direction - a highly conservative worst-case estimate. Adding in quadrature (√(Δx² + Δy²)) is statistically more realistic: it treats uncertainties as independent random variables and gives the standard deviation of their sum. The quadrature result is always smaller than or equal to the direct sum. For two equal uncertainties, quadrature gives a result √2 times smaller than direct addition.
How many significant figures should I keep in the uncertainty?+
By convention, uncertainties are rounded to 1 or 2 significant figures. The central value is then rounded to match the same decimal place as the uncertainty. For example, if your calculation gives 12.347 ± 0.0831, round the uncertainty to 0.08, giving the final result as 12.35 ± 0.08. Never report more decimal places in the central value than in the uncertainty.
When should I use the power rule for uncertainty propagation?+
Use the power rule (Δz/z = n · |Δx/x|) when your quantity is raised to a power - for example, volume = (4/3)πr³ has exponent n = 3, so Δv/v = 3 · Δr/r. The rule also applies to roots: √x has n = 0.5, so Δ(√x) = 0.5 · Δx/√x. For negative exponents (e.g., x⁻²), use the absolute value of the exponent.
What is the half-range method for single measurements?+
When you make only one measurement and cannot compute a standard deviation, you estimate uncertainty from the instrument's resolution. The half-range uncertainty is half the smallest scale division. For a digital scale reading to 0.1 g, the half-range uncertainty is ±0.05 g. For an analogue ruler with 1 mm divisions, it is ±0.5 mm. This represents the maximum plausible error in reading the instrument.
How does the σ/√n method work for repeated measurements?+
If you make n independent measurements of the same quantity and they vary due to random error, the best estimate of the true value is the mean (x̄), and the uncertainty in that mean is the standard error: σ/√n, where σ is the sample standard deviation. Making more repeated measurements reduces the uncertainty proportionally to 1/√n - doubling the number of measurements reduces uncertainty by a factor of √2 ≈ 1.41.
Can I use this calculator for more than two variables?+
The addition/subtraction and multiplication/division modes handle exactly two variables. For three or more variables, apply the rules iteratively: first propagate A and B to get C with uncertainty ΔC, then propagate C and D, and so on. Alternatively, for sums of n independent variables, the general formula is Δz = √(Δx₁² + Δx₂² + ... + Δxₙ²), and for products it is Δz/z = √((Δx₁/x₁)² + (Δx₂/x₂)² + ... + (Δxₙ/xₙ)²).
What is relative uncertainty and how is it used in practice?+
Relative uncertainty (δx = Δx/x) is a dimensionless ratio that expresses uncertainty as a fraction of the measurement. It is particularly useful when comparing the precision of different instruments or measurements. A relative uncertainty below 1% is generally considered high precision; above 10% is considered low precision. In laboratory reports, relative uncertainty is often quoted as a percentage error to communicate the quality of a measurement.