Qualify accuracy with the age buffer

What the unit asks

A system asked whether someone is over eighteen may answer yes only when it estimates over twenty-five. The gap reduces false positives near the line and changes who is sent to a second check. The draft also proposes measuring the distance between a wrong answer and the missed threshold, so that a seventeen-year-old wrongly passed is distinguished from a thirteen-year-old.

What good practice looks like

Record the legal threshold, the challenge age and the secondary route for the uncertain band. Report the false positive absolute error and the mean of it over the false positives, if the sample supports it.

Evidence an assessor asks for

What a good answer sounds like

A good answer to P3-05 names The buffering policy, with both ages and the route for the band between them., dates it, and says what it does not cover. A policy or a promise is not an answer. If P3-05 has no artefact, the row is not met and the absence is the finding.

Where comparisons commonly fail

• The buffer is applied in the live service but not stated in the report.
• The test uses the legal threshold while the service uses a challenge age, so the two figures describe different decisions.
• The near-threshold cases are too few to analyse and the report does not say so.

Questions to put to the supplier

1. What age did the system actually test against?
2. How wrong were the wrong answers, and who did they let through?

How this is used in the kit

P3-05 becomes one row of the comparison register, with an owner, a status and the evidence named above. A status of met asks for the artefact and its date, and the free check uses the same row as an evidence question, so the page and the register share one vocabulary.