Measurement uncertainty and the GUM

A result such as a mass fraction, length or voltage does not fully describe a measurement without its conditions and associated uncertainty. The measurand is the quantity intended to be measured; the measurement model connects it to observations, corrections and other inputs. Uncertainty is different from a known error and from the tolerance assigned to a product. Confusing these concepts can make a technically valid measurement appear to prove more than it does, especially near an acceptance boundary.

The GUM is a body of metrological guidance maintained through the Joint Committee for Guides in Metrology, with publications hosted by BIPM. The official catalogue distinguishes the original guide, supplements, introductory material and related guides. These documents address evaluation and expression of measurement uncertainty. They are not product specifications, and their existence does not determine whether a particular result passes a legal or contractual limit. That decision needs the applicable requirement and decision process.

The JCGM catalogue lists JCGM 100:2008, its 2026 Amendment 1, GUM-1:2023 and the measurement-model guide GUM-6:2020. The original year remains part of a document’s identity even when an amendment is added. Publication year is recorded at the granularity provided by that catalogue rather than supplemented with an assumed day. Official JCGM records identify the documents selected for this reference. A related guide is not presumed to supersede the original unless the publisher’s record establishes that relationship.

In a dimensional inspection, an instrument indication may need corrections and information about temperature, resolution, repeatability and reference standards. The uncertainty budget makes the important contributions visible in relation to the defined measurement. A different setup, operator arrangement or sample condition can alter the budget. Treating a calibration certificate’s uncertainty as the uncertainty of every later measurement overlooks the rest of the process. The working record connects the result to the actual model and conditions under which the item was measured.

The GUM is widely useful technical guidance, but a specific regulator, contract or method may prescribe how uncertainty is evaluated or used. NIST TN 1297 provides guidance for NIST measurement results and is a valuable U.S. reference; it is not a blanket licensing rule for laboratories. A conformity decision near a limit can also involve an agreed decision rule. Measurement uncertainty and the rule for accepting or rejecting an item are related but separate matters.

Examine the quantity, units, measurement conditions and reported uncertainty together. Is the uncertainty standard or expanded? If expanded, is the coverage factor or the basis of the coverage statement given? Can the reader understand how the result and uncertainty apply to the identified sample or instrument? NIST’s reporting guidance emphasizes sufficient information to interpret the statement. For a decision, also identify the limit, agreed rule and responsibility for applying it.

A smaller numerical uncertainty is not automatically evidence of a better decision if the wrong quantity was measured or the sample was unsuitable. Likewise, two results cannot be compared responsibly without considering their definitions, conditions and associated information. This overview explains those relationships without reproducing calculation procedures or promising a general pass/fail algorithm. A defensible budget requires technical knowledge of the measurement process and a justified treatment of contributions, correlations and model limitations.

Testing and calibration laboratories explains how measurement capability appears in a scope and report. Conformity assessment vocabulary distinguishes a measurement result from a conformity statement. The test evidence assessment helps organize questions about the measurand, method, result and decision rule. Read these together when a reported value is close to a specification limit or when results from different laboratories appear inconsistent.

Sources and bibliographic records were checked on 19 October 2026. The BIPM JCGM publication catalogue provides official document links. The NIST guidance is openly available; neither source is presented here as a universal legal rule for every test report. The explanatory examples on this page are original; no third-party guide is mirrored or offered as an NCQM publication. For the mathematical method, use the official document appropriate to the measurement. Submit a correction request identifying the guide, source and statement if a bibliographic or technical issue needs review.