ISO 17025 measurement uncertainty is an important part of demonstrating that a testing or calibration laboratory produces technically reliable results. It helps laboratories quantify the range of values reasonably associated with a measurement and understand how factors such as equipment, methods, environmental conditions, and sample handling affect results.
Under ISO/IEC 17025:2017, Clause 7.6 addresses the evaluation of measurement uncertainty. Laboratories performing calibrations must evaluate uncertainty for their calibration activities, including sampling where relevant. For testing, laboratories must evaluate uncertainty when required by the standard, taking into account significant contributions to the result.
The challenge is not simply calculating a number. Laboratories need a documented, technically justified approach that matches their methods, equipment, intended use, and accreditation scope. For a broader view of the standard, see our ISO 17025 compliance guide.
What Is Measurement Uncertainty in ISO 17025?
Measurement uncertainty describes the dispersion of values that could reasonably be attributed to a measured quantity, based on the information available.
No measurement is perfectly exact. Even when a laboratory follows an approved method and uses calibrated equipment, several factors can influence the reported result.
For example, a laboratory may report a mass measurement of 100.00 g with an expanded uncertainty of ±0.05 g. This indicates that the reported result is accompanied by a quantified uncertainty under the stated measurement conditions and coverage assumptions. It does not mean that the true value is guaranteed to fall within that interval in every case.
Measurement uncertainty helps laboratories:
- Understand the reliability of reported results.
- Identify major sources of variation in a measurement process.
- Assess whether a method is fit for its intended purpose.
- Support comparisons between measurement results.
- Make appropriate decisions when evaluating results against specifications.
The principles for evaluating and expressing uncertainty are described in the Guide to the Expression of Uncertainty in Measurement (GUM), published through the Joint Committee for Guides in Metrology.
Measurement Uncertainty vs. Measurement Error
These terms are related but not interchangeable.
- Measurement error is the difference between a measured value and a reference or true value, where that value is known or can be established.
- Measurement uncertainty characterizes the dispersion of values reasonably attributable to the measurand, based on available information.
Uncertainty is not simply an estimate of how wrong a result might be. It is a quantified part of the measurement process that helps users interpret the result appropriately.
ISO 17025 Measurement Uncertainty Requirements Under Clause 7.6
Clause 7.6 requires laboratories to identify significant contributions to measurement uncertainty and evaluate uncertainty in accordance with the nature of the work. The approach differs depending on whether the laboratory performs testing or calibration.
1. Identify Significant Sources of Uncertainty
Laboratories should consider the factors that can materially affect the result.
Depending on the method, these may include:
- Equipment resolution and calibration uncertainty
- Repeatability and reproducibility
- Environmental conditions, such as temperature or humidity
- Sample preparation and handling
- Method limitations and reference materials
- Operator effects and measurement setup
- Sampling effects, where applicable
The laboratory should justify which contributions are relevant rather than automatically including every conceivable factor.
2. Evaluate Uncertainty for Calibration Activities
For calibration, laboratories must evaluate measurement uncertainty for their calibration activities. The evaluation should reflect the measurement model, relevant uncertainty contributions, and the laboratory's calibration capabilities.
Accreditation bodies may also apply specific policies and guidance to uncertainty statements and calibration and measurement capability. ILAC P14 provides international policy guidance on measurement uncertainty in calibration.
3. Evaluate Uncertainty in Testing
For testing, laboratories must evaluate measurement uncertainty where required by ISO/IEC 17025. If a recognized method specifies limits for the major sources of uncertainty and the way results are expressed, the laboratory must follow the applicable requirements and determine whether the method's instructions are sufficient for its circumstances.
ILAC G17 provides guidance on evaluating and reporting measurement uncertainty in testing.
4. Maintain a Technically Justified Evaluation
The uncertainty evaluation should be appropriate to the measurement process and supported by relevant data, assumptions, calculations, and technical knowledge.
Laboratories should be able to explain why contributions were included or excluded, how values were estimated, and how the final uncertainty was calculated.
How to Calculate ISO 17025 Measurement Uncertainty
A common approach follows the GUM framework. The exact calculation depends on the measurement model, available data, and applicable method or accreditation requirements.
Step 1: Define the Measurand
The measurand is the quantity intended to be measured. Define it clearly, including relevant conditions and the measurement procedure.
For example, the measurand could be the mass of a reference weight under specified conditions or the concentration of a substance in a prepared sample.
Step 2: Identify Uncertainty Contributions
List the significant sources that could affect the measurement result.
For a mass measurement, possible contributions include balance repeatability, display resolution, calibration uncertainty, and environmental effects.
Step 3: Estimate Standard Uncertainty
Express each uncertainty contribution as a standard uncertainty, using appropriate information such as repeated measurements, calibration certificates, specifications, or probability distributions.
Two common approaches are:
- Type A evaluation: Uses statistical analysis of measurement data, such as repeated observations.
- Type B evaluation: Uses other relevant information, such as calibration certificates, manufacturer specifications, reference data, or prior technical knowledge.
Type A and Type B describe methods of evaluating uncertainty, not categories of errors. Both contribute to the overall uncertainty evaluation when relevant.
Step 4: Combine Uncertainty Contributions
Where the measurement model and assumptions allow, independent contributions expressed as standard uncertainties can be combined using a root-sum-of-squares calculation.
Here, (u_c(y)) is the combined standard uncertainty and ui represents each standard uncertainty contribution expressed on the appropriate scale.
If contributions are correlated, covariance terms may need to be included. The calculation should follow the measurement model rather than assuming that all contributions are independent.
Step 5: Calculate Expanded Uncertainty Where Appropriate
Expanded uncertainty is calculated by multiplying the combined standard uncertainty by a coverage factor:
Here, U is the expanded uncertainty and k is the coverage factor selected for the intended coverage and statistical assumptions.
A coverage factor of approximately 2 is often used when an interval with roughly 95% coverage is appropriate, but that interpretation is not universal. The laboratory should justify the coverage factor and state it clearly when reporting expanded uncertainty.
Step 6: Document and Review the Evaluation
Record the measurement model, input values, uncertainty contributions, assumptions, calculation method, and resulting uncertainty.
Review the evaluation when relevant changes occur, such as a method revision, equipment replacement, new calibration information, or changes in measurement conditions.
Measurement Uncertainty Example: Building an Uncertainty Budget
An uncertainty budget lists the significant uncertainty contributions and shows how they contribute to the overall result.
Consider a simplified example for a laboratory mass measurement. Assume the following standard uncertainty contributions have already been expressed in compatible units and can be treated as independent.
The combined standard uncertainty is:
If a coverage factor of k = 2 is justified for this example, the expanded uncertainty is:
The result could therefore be expressed in the form:
Measured mass: 100.000 g; expanded uncertainty: 0.036 g, with k = 2.
This is an illustrative calculation, not a universal ISO 17025 acceptance criterion. A real uncertainty budget must use the laboratory's actual measurement model, applicable corrections, distributions, correlations, and other significant contributions.
How to Document and Report Measurement Uncertainty
A technically sound calculation is only useful when the laboratory can explain and support it.
What Should an Uncertainty Budget Include?
An uncertainty budget commonly records:
- The measurand and measurement method
- The measurement model and relevant input quantities
- Significant uncertainty sources
- Estimated values and standard uncertainties
- Probability distributions and sensitivity coefficients where relevant
- Correlations, where applicable
- The combined standard uncertainty
- The coverage factor and expanded uncertainty, where applicable
- Assumptions, references, and approval or review records
The format should reflect the laboratory's method and reporting needs. Not every budget will contain the same number of entries.
How Should Measurement Uncertainty Be Reported?
The reporting approach depends on the type of work, customer requirements, method, and applicable accreditation guidance.
Where uncertainty is reported, the laboratory should make the result and uncertainty understandable, use appropriate units, and state the coverage factor or coverage information when applicable. Any conformity statement should follow the relevant decision rule and applicable requirements.
For calibration laboratories, ILAC P14 provides specific policy guidance on the estimation and statement of uncertainty. For testing laboratories, ILAC G17 addresses uncertainty evaluation and reporting.
How Software Can Support Uncertainty Management
Software can help laboratories maintain consistent calculation records, connect results to equipment and methods, and make technical information easier to retrieve.
When evaluating a digital system, consider whether it supports:
- Controlled calculation templates and documented assumptions
- Links between results, methods, equipment, and calibration records
- Review and approval workflows where required
- Traceable changes to calculations and associated records
- Retrieval of supporting evidence for technical review and assessment
Software does not determine whether an uncertainty evaluation is technically valid. Laboratories remain responsible for choosing appropriate methods, checking calculations, and demonstrating conformity with applicable requirements.
How Scispot Supports Measurement Uncertainty Workflows
Scispot's ISO 17025 LIMS article describes capabilities for maintaining uncertainty budgets and linking uncertainty records to test results, alongside calibration and equipment information.
For laboratories evaluating software, the relevant question is whether the configured system can support their specific uncertainty-management process, including:
- Connecting measurement results to relevant sample, method, and equipment records
- Organizing calibration and maintenance information
- Maintaining supporting data for uncertainty calculations
- Supporting review and retrieval of technical records
- Providing traceability appropriate to the laboratory's intended use
Confirm the exact calculation, versioning, approval, and audit-trail capabilities with Scispot before relying on the platform for a specific compliance use case.
Build a More Consistent Approach to Measurement Uncertainty
Managing ISO 17025 measurement uncertainty requires more than producing a final number. Laboratories need a defensible method, appropriate technical data, documented assumptions, and records that support the result.
Start by reviewing your measurement models, identifying significant uncertainty contributions, and checking that calculations and reporting practices match your laboratory's methods and accreditation requirements.
Scispot can help laboratories organize related sample, equipment, workflow, and technical records in a connected environment. Evaluate the platform against your specific uncertainty-management and compliance needs.
