Measurement Uncertainty: Understanding the Plus-or-Minus on a COA
Share
A laboratory result is an estimate supported by a measurement process. Measurement uncertainty describes the dispersion of values that can reasonably be associated with that result under the stated approach.
Uncertainty is not the same as a mistake
Sample preparation, calibration, instrument performance, and other parts of a method can contribute to uncertainty. Reporting uncertainty does not automatically mean a laboratory has done poor work. It can provide useful context that a bare number leaves out.
Read the accompanying explanation
A report may show an uncertainty value using a plus-or-minus sign. For example, a hypothetical result of 1.00% with an expanded uncertainty of ±0.05 percentage points describes an interval from 0.95% to 1.05% under the stated coverage assumptions.
That interpretation depends on the units and the laboratory's definition. An absolute uncertainty and a relative percentage uncertainty are different, so ask the laboratory which one is being reported.
A confidence level needs context
An expanded uncertainty may be associated with a coverage factor or stated coverage probability. It should not be treated as an unconditional guarantee that every future measurement or every portion of a batch will fall in the interval.
Use uncertainty without changing the result
Do not subtract uncertainty from a reported value and present the smaller number as the laboratory's result. Record the value and its uncertainty together. Decisions made against a threshold depend on the relevant decision rule and testing framework.
USDA's hemp-production guidance discusses uncertainty within its program. That framework does not make a COA a universal legal or safety determination for every finished product.
Sources:
