Measurement uncertainty is a parameter that characterises the range of values within which the true value of a measurement is expected to lie, given a stated level of confidence. Every measurement — no matter how carefully made, no matter how precise the instrument — contains some degree of doubt. Understanding and quantifying that doubt is not a sign of weakness in a measurement system; it is a sign of technical rigour and honesty. ISO/IEC 17025:2017, the international standard for calibration laboratory competence, requires accredited laboratories to determine and report measurement uncertainty for all calibration results.
For quality engineers, production managers, and compliance officers, grasping the practical meaning of measurement uncertainty is essential for interpreting calibration certificates correctly, making fit-for-purpose decisions about instruments, and satisfying auditors who increasingly expect a sophisticated understanding of measurement traceability.
Why Measurement Uncertainty Exists
No physical measurement is perfectly exact. Even the most sophisticated reference standard used by a national metrology institute has some residual uncertainty associated with its realisation of the unit of measurement. That uncertainty propagates downward through the calibration chain to every instrument that is calibrated against it.
Multiple sources contribute to the total measurement uncertainty in a calibration:
- Resolution of the instrument being calibrated: A digital instrument can only display values in discrete steps equal to its resolution. If a thermometer reads to 0.1°C, it cannot resolve differences smaller than that.
- Repeatability: Even under identical conditions, repeated measurements of the same value will differ slightly due to random effects. The standard deviation of these repeated readings contributes to uncertainty.
- Reference standard uncertainty: The calibration certificate of the reference standard used by the laboratory carries its own uncertainty, which must be included.
- Environmental conditions: Temperature, humidity, and vibration in the laboratory at the time of calibration affect the measurement. Even in well-controlled labs, these conditions vary within specified limits.
- Operator effects: Reading parallax, connection quality, and technique differences between operators can introduce variability.
- Instrument drift: Some instruments do not read consistently throughout a measurement session; drift between the start and end of a calibration sequence contributes to uncertainty.
The Guide to the Expression of Uncertainty in Measurement (GUM)
The internationally accepted framework for calculating and expressing measurement uncertainty is the Guide to the Expression of Uncertainty in Measurement, commonly known as the GUM. Published jointly by BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, and OIML, the GUM provides a unified approach that is used by all ISO/IEC 17025 accredited laboratories worldwide, including those accredited under SAC-SINGLAS in Singapore.
The GUM classifies uncertainty contributions into two categories:
Type A evaluation covers components estimated by statistical analysis of repeated measurements. For example, if a calibration technician takes ten repeated readings of the same nominal pressure, the standard deviation of those readings is a Type A contribution to uncertainty.
Type B evaluation covers components estimated by other means — from calibration certificates of reference standards, from manufacturer specifications, from published data tables, from engineering judgement, or from knowledge of the measurement system. The uncertainty of the reference standard used for calibration is a Type B contribution.
Both types are combined using the law of propagation of uncertainty to produce a combined standard uncertainty, which is then multiplied by a coverage factor (typically k = 2 for approximately 95% confidence) to produce the expanded uncertainty that is reported on the calibration certificate.
How to Read Uncertainty on a Calibration Certificate
On a well-prepared calibration certificate from a SAC-SINGLAS accredited laboratory such as Unitest Instruments, measurement uncertainty is stated alongside each measurement result. The reporting format typically follows this pattern:
Expanded uncertainty: ±0.3°C (k = 2, approximately 95% confidence)
This statement means: the laboratory is approximately 95% confident that the true value of the measurement lies within ±0.3°C of the reported result. The coverage factor k = 2 corresponds to a normal distribution with a 95.45% confidence level; some certificates use k = 2.0 for a rounded 95% level.
When interpreting a calibration certificate, check for the following elements:
- Uncertainty is reported for each measurement point, or the worst-case uncertainty across the range is clearly stated
- The coverage factor k is stated
- The confidence level (typically 95%) is stated
- The measurement result and the uncertainty are in the same units
- The uncertainty is clearly distinguished from any stated tolerance or acceptance criterion
If a certificate does not include measurement uncertainty, it does not comply with ISO/IEC 17025 requirements and cannot have been issued by an accredited laboratory operating within its scope. For a complete guide to reading calibration certificates, see our article on how to read and understand a calibration certificate.
Measurement Uncertainty and Fitness for Purpose
The practical application of measurement uncertainty knowledge is the concept of fitness for purpose. For any measurement, there is a required accuracy — typically expressed as a tolerance or a specification limit. For a measurement to be meaningful, the measurement uncertainty must be sufficiently small relative to the tolerance.
A commonly used guideline is the 4:1 ratio rule, also called the test accuracy ratio (TAR): the measurement uncertainty of the calibration standard should be no more than one-quarter (25%) of the tolerance being verified. If a temperature transmitter must be accurate to ±2°C to satisfy a process requirement, the calibration equipment should have an uncertainty no greater than ±0.5°C.
When the uncertainty is comparable to or larger than the tolerance, the calibration result cannot reliably confirm whether the instrument is within specification. In such cases, the laboratory should flag the situation and the user must decide whether to accept the risk, use a more accurate reference standard, or reduce the tolerance requirement.
This consideration is especially important for high-precision applications — semiconductor fabrication environments, pharmaceutical clean rooms, metrology-grade electrical measurements — where the chain of traceability must be carefully managed to maintain a defensible accuracy ratio at every level. Unitest Instruments' calibration laboratory is equipped with reference standards that maintain appropriate accuracy ratios across its eight accredited disciplines. Contact us via our calibration services page for guidance on specific requirements.
Measurement Uncertainty in Singapore Regulatory Contexts
Several Singapore regulatory frameworks make direct or implied reference to measurement uncertainty requirements:
HSA GMP for pharmaceutical manufacturers: The Health Sciences Authority's GMP guidelines require that measuring equipment used in manufacturing and quality control be calibrated to a known accuracy and that the calibration be traceable. In practice, this means certificates must include uncertainty statements, and the uncertainty must be appropriate for the measurement's intended use.
NEA environmental monitoring: The National Environment Agency requires accredited test reports for emissions monitoring and certain environmental sampling activities. Measurement uncertainty is an inherent part of accredited reporting under ISO/IEC 17025.
PUB water quality monitoring: PUB's requirements for water quality monitoring by licensees reference standards-based measurement, which in the context of ISO/IEC 17025 accredited laboratories includes uncertainty reporting.
ISO 9001 quality management: ISO 9001:2015 clause 7.1.5 requires organisations to ensure that monitoring and measuring resources are fit for purpose, which in a metrologically rigorous interpretation includes understanding measurement uncertainty relative to the tolerance of the measurement.
IATF 16949 and automotive supply chain: Automotive manufacturers' quality systems typically specify explicit requirements for measurement system analysis (MSA), which is a statistical approach closely related to uncertainty analysis. Calibration certificates with stated uncertainty support MSA compliance.
Common Misunderstandings About Measurement Uncertainty
Several misunderstandings about measurement uncertainty frequently arise in industrial and commercial settings.
Misunderstanding 1: A smaller uncertainty always means a better calibration. Uncertainty depends on the reference standard used, the calibration method, and the instrument being calibrated. A calibration performed with a highly accurate reference standard but with poor repeatability of the instrument under test will still report a large uncertainty due to the instrument's own performance. The uncertainty reported on a calibration certificate reflects the performance of the entire calibration process, not just the quality of the reference standard.
Misunderstanding 2: Measurement uncertainty is the same as instrument accuracy. Instrument accuracy (as specified by the manufacturer) is a specification of the instrument's expected performance under defined conditions. Measurement uncertainty as reported on a calibration certificate is an estimate of the doubt in the specific measurement result obtained during that calibration, under the actual conditions that prevailed. They are related but not identical.
Misunderstanding 3: Only laboratories with the most advanced equipment can achieve small uncertainties. While reference standard quality is a major contributor to uncertainty, good laboratory practice — controlled environment, experienced technicians, well-documented procedures, and appropriate measurement methods — makes a significant difference. SAC-SINGLAS assessment evaluates the complete uncertainty budget, not just the equipment on the shelf.
For context on how calibration intervals interact with measurement uncertainty over the life of an instrument, see our article on how often to calibrate instruments.
Requesting and Evaluating Uncertainty Information
When selecting a calibration laboratory or reviewing a calibration certificate, consider the following questions to evaluate the quality of uncertainty reporting:
- Is an expanded uncertainty stated for each measurement result, or at minimum for the worst-case point across the calibration range?
- Is the coverage factor k stated, and is the confidence level specified?
- Is the uncertainty small enough (relative to the instrument's intended tolerance) to make the calibration result meaningful?
- Has the laboratory demonstrated that its uncertainty budget complies with accreditation body requirements through proficiency testing?
- Is the laboratory accredited by SAC-SINGLAS or an equivalent ILAC-MRA signatory body, providing independent assurance that the uncertainty evaluation has been assessed?
Unitest Instruments provides full uncertainty statements on all calibration certificates issued under its SAC-SINGLAS accreditation LA-2023-0845-C. Our technical team is available to explain uncertainty values and their implications for your specific measurement application. To discuss your calibration requirements, contact our team or visit our calibration services page.