ISO 13528-2022 PDF

Full title and description

Statistical methods for use in proficiency testing by interlaboratory comparison — ISO 13528:2022 provides detailed, practical statistical procedures and recommendations for designers and providers of proficiency testing (PT) schemes, for analysing PT data, and for interpreting participant performance. It covers quantitative and qualitative (nominal) results and includes robust estimators, outlier handling, scoring statistics and multi‑round assessments used by laboratories, accreditation bodies and PT providers.

Abstract

This International Standard specifies statistical methods to be used when designing proficiency testing schemes and when analysing the data they produce. It gives guidance on selection of an assigned value, on estimating the standard deviation for proficiency assessment, on detection and treatment of outliers, and on calculating participant performance indicators (z‑scores and variants such as z′). The standard also addresses interpretation of results by participants and accreditation bodies and includes methods applicable where results are quantitative measurements or qualitative observations.

General information

• Status: Published (current edition)
• Publication date: 18 August 2022
• Publisher: International Organization for Standardization (ISO)
• ICS / categories: 03.120.30 (Applications of statistical methods)
• Edition / version: Edition 3 (2022)
• Number of pages: 93

Scope

ISO 13528:2022 applies to proficiency testing where participant results are either quantitative measurements or qualitative (nominal) observations. It provides methods for: designing PT schemes; determining an appropriate assigned value; estimating the standard deviation for proficiency assessment (the criterion used to scale performance scores); detecting and handling outliers; calculating performance statistics (z‑score and its variants); and interpreting PT outcomes for participants, PT providers and accreditation bodies. The document also notes how the procedures can be adapted for assessment of expert judgment when results can be compared objectively with a reference or consensus value.

Key topics and requirements

• Selection and estimation of the assigned value (reference value or consensus statistic) and consideration of its uncertainty.
• Methods to estimate the standard deviation for proficiency assessment (sigma_pt) and guidance on choosing an appropriate criterion for different schemes.
• Robust estimators and algorithms (for example Algorithm A and related robust location/scale methods) to reduce the influence of atypical results.
• Outlier detection and treatment (recommended statistical tests and practical approaches, e.g., Cochran and Grubbs tests as part of conventional procedures, with guidance on robust alternatives).
• Performance scores and interpretation: calculation and use of z‑scores, z′‑scores (z‑prime, which accounts for uncertainty of the assigned value), zone/plot presentations and multi‑round summary statistics (for example RLP and zone plots).
• Approaches for qualitative (nominal) PT items and for assessing categorical/inspection results where comparisons with assigned or consensus values are possible.
• Guidance on multi‑round schemes, aggregated performance metrics, and presentation of results to participants and accreditation bodies.
• Recommended reporting practices, including communicating when assigned‑value uncertainty is not negligible and how that affects participant evaluations.

Typical use and users

Primary users are proficiency testing providers and scheme designers who need standardized, defensible statistical methods for scheme design, data analysis and participant scoring. Secondary users include testing and calibration laboratories (to interpret PT reports), accreditation bodies (for assessment and oversight), statisticians working in conformity assessment, reference‑material producers and regulatory or conformity assessment stakeholders who rely on interlaboratory comparison results.

Related standards

Standards commonly referenced alongside ISO 13528 include ISO/IEC 17043 (conformity assessment — competence of proficiency testing providers), ISO/IEC 17025 (competence of testing and calibration laboratories), the ISO 5725 series (accuracy — trueness and precision of measurement methods), ISO 17034 (reference material producers) and earlier editions of ISO 13528 (for example ISO 13528:2015, now superseded). Guidance documents and harmonized protocols from professional bodies (IUPAC/AOAC harmonized approaches to PT scoring) are also frequently used together with ISO 13528.

Keywords

proficiency testing; interlaboratory comparison; statistical methods; assigned value; sigma_pt; z‑score; z′‑score; robust statistics; Algorithm A; outlier detection; Cochran; Grubbs; multi‑round assessment; accreditation.

FAQ

Q: What is this standard?

A: ISO 13528:2022 is an International Standard that specifies statistical methods for designing, analysing and interpreting proficiency testing (PT) schemes based on interlaboratory comparisons.

Q: What does it cover?

A: It covers selection/estimation of assigned values, estimation of the standard deviation for proficiency assessment, robust estimators and outlier handling, calculation of participant performance scores (z‑scores and variants), multi‑round analysis and guidance on interpreting PT results for quantitative and qualitative data.

Q: Who typically uses it?

A: Proficiency testing providers, scheme designers, statisticians working in interlaboratory studies, testing and calibration laboratories (to interpret PT reports), and accreditation bodies involved in oversight of PT activities.

Q: Is it current or superseded?

A: ISO 13528:2022 is the current (third) edition published in August 2022 and supersedes ISO 13528:2015.

Q: Is it part of a series?

A: It sits alongside related ISO standards used in conformity assessment and measurement quality, notably ISO/IEC 17043 (proficiency testing providers) and the ISO 5725 series on accuracy of measurement methods. Previous editions of ISO 13528 form the historical sequence (2005, 2015, 2022).

Q: What are the key keywords?

A: Assigned value, sigma_pt (standard deviation for proficiency assessment), z‑score, z′‑score (z‑prime), robust statistics, Algorithm A, outlier detection, proficiency testing, interlaboratory comparison, accreditation.