§8.3 · BUILDING · 5 MIN READ · EDITION 1.0
Coverage, missed findings and false findings
These three measures, read together, describe what a tool actually does on real evidence better than any single accuracy number.
A high accuracy number can hide a dangerous pattern: perfect performance on easy cases and silent failure on the rare, consequential ones that correlated-error testing is designed to expose.
Report all three measures with the reference-set composition so a reader can judge whether the result generalises to their own evidence.
PUT THIS INTO PRACTICE
- 01Report coverage, missed and false findings together.
- 02Publish the reference-set composition alongside results.
- 03Re-test after any material configuration change.
RELATED INSTRUMENTS
CITE THIS SECTION
§8.3 · Edition 1.0 · Coverage, missed findings and false findings