What it measures
The ALiF-SAQ asks people to rate their own experience and confidence across 48 core items covering the sixteen sub-competencies, with up to 8 additional items for each selected professional role. It takes about 20 to 30 minutes, or up to 40 minutes with several role modules. Scores are averaged by sub-competency and component and placed on the three ALiF levels.
What it is for
The results are used to recommend a learning pathway: where to start, which modules to take and which to skip. The questionnaire is designed for development and diagnosis, not for credentialing.
What it does not measure
It measures how competent people believe they are, not how they perform. Self-ratings can be inflated, especially at lower levels of competence, and are affected by how people want to be seen. The SAQ does not assess performance in real AI use, does not produce safety assessments, and does not replace observation or performance-based evaluation.
Known limitations
Some sub-competencies are measured by only one or two items, which limits their reliability. The questionnaire has been piloted in health care and higher education in one country; its properties in other regions, languages and professions have not yet been established.
Validation
- Phase 1 (under way): content validity with an expert panel, cognitive testing, a pilot and exploratory factor analysis.
- Phase 2 (planned): confirmatory factor analysis, testing whether the questionnaire works the same way across roles, and links between scores and actual AI use.
To protect the validation, the item wording is not published on this site. Coordinators receive the questionnaire and administrator guide from the ALiF core team.