✦ScienexySCIENCE KNOWLEDGE LIBRARY
reference · SCI-REF-003-088

Uncertainty Budgets and Reporting Measurement Results

An uncertainty budget is a structured summary of the uncertainty contributions associated with a measurement result. It documents input quantities, uncertainty sources, evaluation methods, probability distributions, sensitivity coefficients, correlations and their contributions to combined uncertainty.

For uncorrelated input quantities, u_c²(y) = Σ[cᵢu(xᵢ)]². Expanded uncertainty is U = k u_c(y). For correlated inputs, covariance contributions must also be included.

Explanation

A measurement uncertainty budget organizes the information required to calculate and report the uncertainty of a measurement result. It begins with a clearly defined measurand and measurement model. Each relevant input quantity is identified, and its uncertainty is evaluated using Type A or Type B methods. The uncertainty is converted into standard uncertainty and multiplied by the appropriate sensitivity coefficient to determine its contribution to the output uncertainty. Correlations between input quantities are included through covariance terms when relevant. For uncorrelated contributions, their squared standard contributions are added and the square root gives the combined standard uncertainty. An uncertainty budget makes the evaluation transparent and allows the largest contributions to be identified. It can also reveal unnecessary complexity, duplicated sources or missing information. Reporting should include the measurement result, unit, combined or expanded uncertainty, coverage factor when applicable, measurement method and relevant conditions. A calibration or laboratory report may additionally include traceability information, reference standards, calibration dates and the uncertainty evaluation method. The number of displayed digits should be consistent with the uncertainty and the reporting requirements. Raw observations, corrections and calculation records should be retained so that the result can be reviewed and reproduced.

Variables

SymbolMeaningSi Unit

u_c(y)

Combined standard uncertainty of the output quantity

Same unit as y

u(xᵢ)

Standard uncertainty of input quantity i

Same unit as xᵢ

cᵢ

Sensitivity coefficient of input quantity i

Unit of y divided by unit of xᵢ

U

Expanded uncertainty

Same unit as y

k

Coverage factor

Dimensionless

Example or application

A worked example has not yet been added.

Definition

An uncertainty budget is a documented account of the uncertainty sources, their standard uncertainty values, sensitivity coefficients, correlations and contributions used to evaluate the combined standard uncertainty of a measurement result.

Purpose

To document the complete uncertainty evaluation, show how individual sources influence the result and support transparent, traceable and reproducible measurement reporting.

Importance

Uncertainty budgets are important in calibration laboratories, accredited testing, manufacturing, research, instrument validation and conformity assessment. They provide evidence that reported uncertainty values are based on a structured evaluation.

Uncertainty Budget Components

ComponentDescription

Measurement model

The mathematical relationship between the measurand and the relevant input quantities.

Input quantity

A measured, calculated or reference quantity used in the measurement model.

Uncertainty source

The physical, instrumental, environmental, methodological or statistical influence associated with an input.

Evaluation method

Type A or Type B method used to estimate standard uncertainty.

Probability distribution

The distribution assumed or supported by the available information.

Standard uncertainty

The uncertainty contribution expressed in standard-deviation form.

Sensitivity coefficient

The derivative or equivalent coefficient describing the influence of the input on the output.

Uncertainty contribution

The standard uncertainty multiplied by its sensitivity coefficient.

Correlation information

Covariance or correlation information used when input quantities are statistically related.

Combined uncertainty

The resulting standard uncertainty after all contributions and correlations are considered.

Uncertainty Budget Development

  • Define the measurand and measurement conditions.

  • Develop the measurement model.

  • List every significant input quantity.

  • Identify possible uncertainty sources for each input.

  • Evaluate Type A and Type B contributions.

  • Convert each contribution into standard uncertainty.

  • Calculate sensitivity coefficients.

  • Determine whether input quantities are correlated.

  • Calculate the combined standard uncertainty.

  • Select an appropriate coverage factor if expanded uncertainty is required.

  • Review the budget for missing, duplicated or insignificant contributions.

  • Document the final result and all relevant assumptions.

Illustrative Budget

Measurement Model

y = x₁ + x₂

Assumption

The two input quantities are uncorrelated and their sensitivity coefficients are both 1.

Contributions

SourceStandard UncertaintySensitivity CoefficientStandard Contribution

Input 1

0.03 mm

1

0.03 mm

Input 2

0.04 mm

1

0.04 mm

Calculation

u_c(y) = √(0.03² + 0.04²) = 0.05 mm

Result

The combined standard uncertainty is 0.05 mm.

Interpretation

The combined uncertainty is calculated from the squared standard contributions, not by directly adding their magnitudes.

Reporting Measurement Results

Essential Information

  • Measurement result.

  • Measurement unit.

  • Combined or expanded uncertainty, as appropriate.

  • Coverage factor when expanded uncertainty is reported.

  • Measurement method and relevant conditions.

  • Identification of the measurand.

Additional Information Where Relevant

  • Calibration status and traceability.

  • Reference standards and materials.

  • Measurement date and environmental conditions.

  • Uncertainty evaluation procedure.

  • Distributional assumptions and degrees of freedom.

  • Correction factors applied.

  • Limitations and validity range.

Example

Mass = (125.43 ± 0.05) g, where the stated uncertainty is expanded uncertainty with coverage factor k = 2.

Interpretation

The report should make clear that ±0.05 g is expanded uncertainty and that the coverage interpretation depends on the uncertainty evaluation.

Uncertainty Budget Review

  • Check that all important uncertainty sources have been considered.

  • Confirm that the same source has not been counted twice.

  • Verify the units of all standard contributions.

  • Check sensitivity coefficients and their signs where relevant.

  • Review correlations and assumptions of independence.

  • Confirm that the selected distributions are justified.

  • Check calculations and rounding.

  • Verify that the reported coverage factor is appropriate.

Limitations

  • An uncertainty budget is only as complete as the information and model used to construct it.

  • An omitted source cannot be accounted for by accurate arithmetic.

  • Correlations can make simple root-sum-of-squares calculations inappropriate.

  • Uncertainty budgets may require revision when instruments, methods or environmental conditions change.

  • A detailed budget does not guarantee that the measurement process is free from unrecognized bias.

Reference

  • JCGM 100:2008, Evaluation of Measurement Data — Guide to the Expression of Uncertainty in Measurement (GUM).

  • NIST Technical Note 1297, Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results.

  • BIPM, The International System of Units (SI Brochure), 9th edition.

Related Topic

  • SCI-REF-003-081

  • SCI-REF-003-082

  • SCI-REF-003-086

  • SCI-REF-003-087

  • SCI-REF-003-079

Sources and verification

Sources are shown as provided in the project data. Not every source has yet been independently verified. Please report errors or missing citations.

  • Source verification pending; check an authoritative textbook, university, standards body, or peer-reviewed source.

Suggest a correction or source · More in Units & Measurement