ISO 5725-4:2020

ISO 5725-4:2020 - Overview

ISO 5725-4:2020 provides basic methods for determining the trueness (bias) of a standardized measurement method. It describes practical, statistically‑based procedures to estimate the bias of a measurement method (via interlaboratory experiments) and the laboratory bias for a single laboratory. The standard applies to methods that return a single numerical result on a continuous scale and only when an accepted reference value (substitute for the true value) can be established (e.g., reference materials, measurement standards, reference methods, or prepared known samples).

Key topics and technical requirements

• Scope and applicability
  • Applies to standardized measurement methods producing one continuous value per result.
  • Not applicable where measurement bias depends on levels of other properties (no interference modelling).
• Design of experiments
  • Procedures for interlaboratory experiments to estimate method bias and for single‑laboratory tests to estimate laboratory bias.
  • Guidance on layout, required number of laboratories and measurements, and experimental objectives.
• Accepted reference values
  • Requirements and approaches to assign an accepted reference value and to account for its measurement uncertainty.
  • Examples: reference materials, standards, validated alternative methods, or prepared samples.
• Statistical evaluation
  • Estimation of bias and testing for statistical significance (with standard significance level α = 0.05).
  • Checks on precision (within‑laboratory repeatability and between‑laboratory reproducibility) and consistency tests (e.g., Mandel, Grubbs tests referenced in the standard).
  • Reporting requirements and the role of a statistical expert and decision panel.
• Reporting and decisions
  • How results are reported to panels and the decisions that follow (acceptance of bias, maximum undetected bias, corrective actions).

Practical applications and users

ISO 5725-4 is used by:

• Accreditation bodies and proficiency testing providers setting performance criteria for methods.
• Testing and calibration laboratories wanting to estimate or monitor laboratory bias.
• Standards committees and method developers validating the trueness of standardized methods.
• Metrologists, QA/QC managers, and statisticians involved in uncertainty evaluation and method validation.

Practical outcomes include validated estimates of method bias, documented laboratory bias checks, improved measurement accuracy (trueness and precision), and defensible uncertainty statements.

Related standards

• ISO 5725-1: General principles and definitions (trueness and precision)
• ISO 5725-2: Repeatability and reproducibility experiments
• ISO 5725-6: Comparison of trueness of two methods
• ISO Guide 30: Reference materials - terms and definitions
• ISO 3534‑1 / ISO 3534‑2: Statistical vocabulary and symbols

Keywords: ISO 5725-4, trueness, bias, laboratory bias, interlaboratory experiment, measurement method, accepted reference value, reference material, measurement uncertainty, accuracy of measurement.