ISO/IEC 17025 applies to laboratories performing testing and calibration activities. Clauses 7.6 and 7.7 work together to ensure that laboratory results are both technically valid and consistently reliable.
Clause 7.6 focuses on measurement uncertainty, requiring laboratories to understand and evaluate factors that influence measurement results. Clause 7.7 addresses ensuring the validity of results, requiring laboratories to monitor performance over time using both internal and external checks.
Together, these clauses support confidence—in laboratory data, laboratory processes, and laboratory competence.
Clause 7.6: Measurement Uncertainty
What Is Measurement Uncertainty?
Measurement uncertainty is an expression of the doubt that exists in any measurement result. No measurement is exact, and ISO/IEC 17025 requires laboratories to understand what affects their results and how significant those effects are.
Clause 7.6 requires laboratories to:
- Identify significant contributors to measurement uncertainty
- Evaluate uncertainty using appropriate methods
- Apply uncertainty evaluation consistently with laboratory activities
The intent is not perfection but understanding.
Identifying Contributors to Uncertainty
Uncertainty contributors can arise from many sources, including:
- Reference standards
- Equipment resolution
- Repeatability of measurements
- Environmental conditions such as temperature or humidity
- Equipment drift over time
- Sampling, when applicable
Not every possible contributor must be included—only those that significantly affect the measurement result. The evaluation should be appropriate to the measurement being carried out.
Type A and Type B Uncertainty
Uncertainty contributors are commonly categorized as Type A or Type B.
- Type A uncertainty is evaluated using statistical methods, such as repeated measurements and calculation of standard deviation.
- Type B uncertainty is evaluated using non-statistical information, including calibration certificates, manufacturer specifications, reference materials, published data, or professional judgment.
Most uncertainty budgets include a combination of both. The key is not the classification, but whether the contributor is technically justified and properly quantified.
Uncertainty Distributions
To estimate uncertainty, laboratories apply probability distributions to contributors.
Common distributions include:
- Normal distribution, often used for statistically derived (Type A) uncertainty
- Rectangular (uniform) distribution, commonly used for specifications or resolution limits
- Triangular or U-shaped distributions, used less frequently in specific cases
When uncertainty exists about which distribution applies, a rectangular distribution is often a conservative and acceptable choice when justified.
Calibration vs. Testing Laboratories
Clause 7.6 applies differently depending on laboratory activities.
- Calibration laboratories must evaluate measurement uncertainty for all calibrations, including in-house calibrations supporting testing.
- Testing laboratories must evaluate uncertainty when the test method allows. If a rigorous evaluation is not possible, a reasonable estimation based on theory or experience is required.
When recognized test methods specify how uncertainty is addressed or how results are reported, following the method correctly may satisfy the requirement.
Clause 7.7: Ensuring the Validity of Results
Clause 7.7 focuses on monitoring laboratory performance to ensure results remain valid over time. This monitoring must be planned, implemented, recorded, and reviewed.
Internal Monitoring Activities
Laboratories must use internal controls to detect trends or issues before incorrect results are reported.
Examples include:
- Use of reference materials
- Functional and intermediate equipment checks
- Replicate testing
- Control charts
- Review of results prior to release
Where practical, statistical techniques should be used, and records must demonstrate that monitoring activities are performed and reviewed.
External Monitoring and Proficiency Testing
In addition to internal monitoring, laboratories must participate in external comparisons where available and appropriate.
This includes:
- Proficiency testing programs
- Interlaboratory comparisons
These activities provide objective evidence of laboratory competence by comparing results with peer laboratories.
When proficiency testing is not available or appropriate, alternative approaches—such as intralaboratory comparison may be used, but they must be justified and documented.
Reviewing Results and Taking Action
Results from internal and external monitoring activities must be analyzed. If results fall outside acceptable criteria, laboratories are required to:
- Investigate the issue
- Implement corrective actions
- Prevent incorrect results from being reported
Monitoring activities are not simply a compliance exercise, they are a critical tool for continual improvement and confidence in laboratory performance.
Why Clauses 7.6 and 7.7 Matter
Measurement uncertainty and result validity are essential to demonstrating technical competence under ISO/IEC 17025. These clauses ensure that laboratories:
- Understand the limitations of their measurements
- Apply consistent and defensible evaluation methods
- Detect issues before they impact customers
- Build trust with accreditation bodies and clients
When applied effectively, Clauses 7.6 and 7.7 strengthen not only compliance, but the overall quality and reliability of laboratory operations.
Contact PJLA
If you have questions about ISO/IEC 17025 accreditation, measurement uncertainty, or how to implement effective systems in your laboratory, Perry Johnson Laboratory Accreditation, Inc. (PJLA) is here to help.
Phone: 1-877-369-5227 (1-877-369-LABS) or (248) 519-2603
Email: [email protected]
Website: https://www.pjlabs.com
Address: Perry Johnson Laboratory Accreditation, Inc., 755 W. Big Beaver Rd., Suite 1325, Troy, Michigan, 48084, USA pjlabs.com
PJLA provides expert accreditation services and support for laboratories seeking compliance with international standards such as ISO/IEC 17025.
