Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO 29992:2018 — Assessment Services for Learning Outcomes — Framework
Quality framework for learning outcome assessment services | Validity, reliability, and fairness in assessment | Assessment lifecycle management
1. Framework for Assessment Services
ISO 29992:2018 specifies requirements for assessment services for learning outcomes. It establishes a framework covering the entire assessment lifecycle — design, development, delivery, and post-delivery — ensuring that assessments are valid, reliable, fair, and transparent. The standard applies to all types of assessment (formative, summative, diagnostic) and delivery modes (face-to-face, technology-based, blended).
Critical Context: Assessment quality directly impacts the credibility of qualifications and certifications. A poorly designed assessment can invalidate an entire learning program, regardless of the quality of instruction. ISO 29992 provides the quality infrastructure needed to ensure assessment integrity.
2. Assessment Lifecycle Requirements
The standard structures requirements around five key processes: assessment framework definition, instrument development, assessment delivery, results processing and reporting, and post-assessment review. Each process includes specific requirements for quality assurance documentation, personnel competence, and continuous improvement. The assessment framework must define the purpose, target population, learning outcomes covered, and the evidence required to demonstrate achievement.
| Process | Key Requirements | Documentation |
|---|---|---|
| Framework definition | Purpose, population, outcomes, evidence requirements | Assessment strategy document |
| Instrument development | Validity evidence, pilot testing, item analysis | Item specifications, pilot report |
| Assessment delivery | Standardized conditions, security, accommodations | Administration manual, security log |
| Results processing | Marking reliability, grading, moderation | Marking scheme, moderation records |
| Post-assessment review | Statistical analysis, stakeholder feedback, improvement | Annual assessment report |
Engineering Insight: The standard’s emphasis on validity evidence mirrors the ISO/IEC 17025 approach to method validation. Assessment instruments must produce results that are “fit for purpose” — a concept directly analogous to measurement uncertainty in testing laboratories.
3. Quality Assurance and Improvement
The standard requires organizations to establish a quality assurance system that includes internal verification of assessment judgments, external moderation where appropriate, and systematic analysis of assessment data to identify improvement opportunities. Assessor competence is explicitly addressed: assessors must have appropriate subject matter knowledge, assessment skills, and ongoing professional development. The standard also addresses candidate rights, including the right to appeal assessment decisions through a formal process.
4. Frequently Asked Questions
Q1: Can ISO 29992 be used for high-stakes examinations?
A: Yes, the standard is designed for all levels of consequence. The rigor of implementation should be proportional to the stakes of the assessment.
A: Yes, the standard is designed for all levels of consequence. The rigor of implementation should be proportional to the stakes of the assessment.
Q2: How does ISO 29992 relate to ISO 21001?
A: ISO 21001 addresses educational organization management systems broadly. ISO 29992 provides specific requirements for the assessment function within such organizations.
A: ISO 21001 addresses educational organization management systems broadly. ISO 29992 provides specific requirements for the assessment function within such organizations.
Q3: Is technology-based assessment covered?
A: Yes, the standard includes requirements for technology-based assessment delivery, including platform reliability, data security, and digital identity verification.
A: Yes, the standard includes requirements for technology-based assessment delivery, including platform reliability, data security, and digital identity verification.
Q4: What is the minimum sample size for pilot testing?
A: The standard recommends a minimum of 30 candidates per pilot administration for statistical analysis, with adaptation for low-volume professional certifications.
A: The standard recommends a minimum of 30 candidates per pilot administration for statistical analysis, with adaptation for low-volume professional certifications.
