ISO 16820-2019 PDF

Full title and description

ISO 16820:2019 — Sensory analysis — Methodology — Sequential analysis. This international standard specifies a sequential statistical procedure for analysing data from forced‑choice sensory discrimination tests (for example triangle, duo‑trio, 3‑AFC, 2‑AFC) in which a decision to stop or continue testing may be made after each trial.

Abstract

The standard defines a sequential testing framework in which, after every trial, the analyst can: stop and declare a perceptible difference, stop and declare no perceptible difference, or continue testing. The sequential method is intended to reach valid decisions with fewer trials than fixed‑sample approaches and is useful for assessing whether formulation, processing, packaging, handling or storage changes produce perceptible differences, and for selecting, training and monitoring assessors. Worked examples and statistical guidance (including use of Thurstonian conversions) are provided.

General information

• Status: Published (International Standard; confirmed as current in ISO review).
• Publication date: October 2019 (Edition 2, 2019).
• Publisher: International Organization for Standardization (ISO).
• ICS / categories: 67.240 — Sensory analysis.
• Edition / version: Edition 2 (2019). Replaces ISO 16820:2004.
• Number of pages: 10 pages in the ISO publication (note: some national adoptions/publisher formats show additional front‑matter and list 15–18 pages).

Scope

ISO 16820:2019 specifies a sequential procedure for the statistical analysis of forced‑choice discrimination test data (triangle, duo‑trio, 3‑AFC, 2‑AFC). It covers construction of decision boundaries, choice of risk parameters (α and β), selection of p0 and p1 probabilities, stopping rules and interpretation of cumulative results. The procedure is applicable at panel and individual assessor levels for determining presence or absence of perceptible differences and for assessor selection, training and monitoring.

Key topics and requirements

• Definition of sequential testing framework for forced‑choice discrimination tests (triangle, duo‑trio, 3‑AFC, 2‑AFC).
• Specification of risk parameters (α, β) and guidance on selecting p0 (chance level) and p1 (performance under a true difference).
• Construction of upper and lower decision boundaries and graphical interpretation (cumulative correct responses vs. trials).
• Sequential stopping rules: stop to accept difference, stop to accept no difference, or continue when results fall between boundaries.
• Statistical conversions and guidance (including use of Thurstonian δ approach) and worked examples in annexes to illustrate application.

Typical use and users

This standard is used by sensory scientists, sensory laboratories, food and beverage R&D teams, quality assurance and production personnel involved in product consistency or shelf‑life studies, and panel managers responsible for selecting, training and monitoring assessors. It is used both for experimental discrimination studies and for operational quality‑control/assessor performance tracking where faster, risk‑controlled decisions are desirable.

Related standards

ISO 16820 is part of the ISO sensory‑analysis family; common related/ referenced standards include ISO 5492 (sensory vocabulary) and other sensory methodology standards such as ISO 6658 (general guidance), ISO 4121, ISO 8586, ISO 8587, ISO 5496 and national adoptions (e.g., BS ISO 16820). Users often apply these together when designing panels, rooms and test methods.

Keywords

ISO 16820, sensory analysis, sequential analysis, discrimination tests, triangle test, duo‑trio, AFC, 2‑AFC, 3‑AFC, alpha risk, beta risk, stopping rules, Thurstonian delta, assessor selection, quality control.

FAQ

Q: What is this standard?

A: ISO 16820:2019 is an international standard that defines a sequential statistical method for analysing forced‑choice sensory discrimination test data (triangle, duo‑trio, 3‑AFC, 2‑AFC).

Q: What does it cover?

A: It covers the choice of risk parameters, construction of decision boundaries, stopping rules (declare difference, declare no difference, or continue), and guidance for converting sensory effect sizes (e.g., via Thurstonian approaches), plus examples illustrating applications.

Q: Who typically uses it?

A: Sensory scientists, panel managers, food/beverage R&D, QA/QC teams and laboratories conducting discrimination testing or monitoring assessor performance.

Q: Is it current or superseded?

A: Edition 2 (2019) replaced the 2004 edition. The ISO record shows the publication was reviewed and confirmed (reviewed status) and remains current as of the ISO review noted in 2025. National adoptions (BS, DIN, etc.) may show their own publication metadata.

Q: Is it part of a series?

A: Yes — it is part of the ISO sensory analysis family of standards produced by ISO/TC 34/SC 12 and is commonly used alongside related ISO methods and vocabulary standards for sensory testing.

Q: What are the key keywords?

A: Sensory analysis, sequential analysis, discrimination tests, triangle, duo‑trio, AFC, alpha risk, beta risk, Thurstonian delta, assessor training/monitoring.