Measurement uncertainty is a parameter expressing the dispersion of values that could reasonably be attributed to what is being measured. Put plainly: every measurement has doubt attached, and uncertainty quantifies it.
Why laboratories must estimate it
ISO/IEC 17025 requires laboratories to identify the contributions to measurement uncertainty and to have a procedure for estimating it. A result reported without any understanding of its uncertainty cannot support a conformity statement, because you cannot say whether a value near a specification limit passes or fails.
Where uncertainty comes from
A budget typically accounts for the reference standard’s own uncertainty, instrument resolution, repeatability across measurements, reproducibility across operators and days, environmental conditions such as temperature and humidity, and the operator’s technique.
Individual contributions are combined and then expanded — most commonly by a coverage factor of k=2, giving roughly 95% confidence.
Decision rules
When a laboratory states that an item conforms to a specification, uncertainty determines what happens near the limit. A decision rule defines how it is handled: whether the uncertainty band must fall entirely inside the limit, or whether a measured value inside the limit suffices. ISO/IEC 17025 requires the rule to be documented and communicated to the customer.
This is a common gap. Laboratories calculate uncertainty diligently and then report pass or fail without stating the rule they applied.
Not the same as calibration
Calibration establishes the relationship between an instrument and a reference. Uncertainty quantifies how much you should trust the result afterwards. You need both.