Measurement Uncertainty

Measurement Uncertainty – have we allowed for variance

Whilst measurement uncertainty is not integral to ISO9001 itself, most quality processes require consideration of data or metrics and by default the uncertainty surrounding the data we analyse may be of interest. Typically, we monitor trends in our numbers, seek explanations as to why variance may exist in data sets, (particularly where they display variance from expected results) and generally find reasons to disregard the outliers.

Also of note in ISO 9001:2015, clause 7.1.5 Measurement Traceability where measurement uncertainty is to some extent controlled through our calibration and verification activities.

We first came across Measurement Uncertainty and the accuracy of numbers too, in flow measurement where it was an important factor in the development of software for measuring the flow of gases and liquids. Flow measurement may be one application of measurement uncertainty but it’s an interesting concept in quality management if we are to fully understand the numbers we are evaluating.

Measurement uncertainty in quality refers to the doubt that exists about the result of a measurement. It’s an important concept in quality control and laboratory testing because it characterizes the dispersion of values that could reasonably be attributed to the measure and—the quantity being measured.

Here are some key points about measurement uncertainty:

• No measurement is perfect: Every measurement has an inherent error, which can be due to various factors such as the precision of the measurement instruments or the method of measurement.
• Components of uncertainty: Measurement uncertainty typically consists of two components—random and systematic errors. Random errors arise from unpredictable variations and can usually be reduced by increasing the number of observations. Systematic errors, on the other hand, cannot be eliminated but can often be reduced by correction if they can be quantified appropriately.
• Reporting with confidence: Measurement uncertainty is reported with a confidence level, typically 95%, which indicates the probability that the true value lies within the uncertainty range.
• ISO standards: ISO/IEC 17025 is an international standard that requires testing and calibration laboratories to estimate and report measurement uncertainty. This ensures the reliability and comparability of laboratory results. ISO9001 is not prescriptive in terms of measurement uncertainty but it may be relevant in some sectors.

Understanding and properly evaluating measurement uncertainty is crucial because it affects the reliability of the decisions made based on the measurement results. Too large or too small an uncertainty can impact the quality and credibility of the results. For most purposes 95% is a good level of accuracy and underpins the importance of assessing statistical trends and monitoring variance.

To discuss aspects of statistical analysis, including sampling and measurement uncertainty, be sure to join one of our QCS CQI IRCA Registered ISO 9001 Lead Auditor training courses.