Overview
ISO 10576:2022 - "Statistical methods - Guidelines for the evaluation of conformity with specified requirements" provides guidance for deciding whether a measured, quantifiable characteristic meets specified limits (limiting values) when the measurement result carries uncertainty. The standard applies whenever measurement uncertainty can be quantified following ISO/IEC Guide‑98‑3 (GUM). It focuses on using the measurement result plus its associated uncertainty to assess conformity and documents recommended reporting and assessment practices. The standard adopts a frequentist perspective and replaces the earlier ISO 10576‑1:2003.
Key topics and requirements
- Scope and definitions: Clear definitions for terms such as limiting values, region of permissible/non‑permissible values, uncertainty interval and confidence level (see ISO 3534 references).
- Specification of limiting values: Recommendations for unambiguous description of the entity and the quantifiable characteristic, requirement for validated or standardized test procedures, and guidance on single (upper or lower) and double limits.
- Measurement uncertainty: Requirement to quantify uncertainty according to the GUM principles; uncertainty is treated as a descriptor of all variation sources, including sampling.
- Conformity assessment procedures:
- One‑stage conformity assessment - assess the measurement result plus uncertainty against limiting values in a single test.
- Two‑stage conformity assessment - an initial stage followed by a second stage to reduce the risk of erroneous decisions (used to balance consumer/supplier risk).
- Use of uncertainty intervals (including confidence intervals) to derive decisions.
- Reporting: How to report measurement results, uncertainty, assurance of conformity or non‑conformity, and handling of inconclusive results (note: the standard does not prescribe actions to take when results are inconclusive).
- Examples and illustrative material: Annex A lists example entities and characteristics; Annex B provides worked examples (updated to include lab‑to‑lab variability).
Applications and users
ISO 10576:2022 is practical for organizations and professionals who must declare or verify conformity in the presence of measurement uncertainty:
- Testing and calibration laboratories
- Quality control and production engineers
- Conformity assessment bodies and third‑party auditors
- Regulatory agencies and environmental/health monitoring programs
- Manufacturers (chemicals, food, pharmaceuticals, materials, oil, etc.)
Use cases include comparing test results to specification limits, generating supplier declarations of conformity, and designing assessment procedures that account for measurement and sampling uncertainty.
Related standards
- ISO/IEC Guide‑98‑3 (GUM) - Guide to the expression of uncertainty in measurement (required)
- ISO 3534‑1 / 3534‑2 - Statistics vocabulary and symbols
- ISO 5725‑1 to ‑6 - Accuracy (bias and precision) of measurement methods and results
- ISO/IEC Guide 98‑4 - Bayesian perspective on uncertainty (complementary)
- ISO/TR 13587 - Comparison of frequentist, Bayesian and fiducial approaches
Keywords: ISO 10576:2022, conformity assessment, measurement uncertainty, GUM, confidence interval, limiting values, statistical methods, one‑stage assessment, two‑stage assessment, specification limits, quality control.