Report continuous measures for an estimated age

What the unit asks

The draft's annex sets out the basic measures for a continuous output: the signed error, its absolute value, the mean and the spread, and the shape of the distribution. The reason is practical. Two systems with the same mean error can behave very differently near the threshold, and a histogram shows the failures a single number conceals.

What good practice looks like

Publish the signed error and the absolute error, a histogram or the quantiles, and a table by age band. Say whether the sample is balanced by age. Where a mean is reported, name the measure and its denominator.

Evidence an assessor asks for

What a good answer sounds like

A good answer to P3-06 names The error table by age band and by demographic group., dates it, and says what it does not cover. A policy or a promise is not an answer. If P3-06 has no artefact, the row is not met and the absence is the finding.

Where comparisons commonly fail

• Only the mean error is published, with no spread and no direction.
• The age bands are so wide that the near-threshold group disappears.
• The chart is published without the underlying counts.

Questions to put to the supplier

1. What does the tail of the error distribution look like?
2. Which age band carries the largest absolute error?

How this is used in the kit

P3-06 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.