Precision measurement: full census of 176 candidates, 2 defects fixed
A full-census (not sampled) precision measurement by independent judges, refusing to compare a stricter new figure to a stale, easier one.
A full-census (not sampled) precision measurement by independent judges, plus two real extraction defects found and fixed with regression tests observed failing before the fix. An earlier precision figure (87.5%) was computed under a looser standard than a later one (45.5%), and the two were at risk of being compared as if they meant the same thing. Evidence: PRECISION-POST-6510.md, which grades all 176 candidates individually against a stated standard.
Technologies
- Full-census grading
- Independent judge panel
- Regression-tested defect correction
Applicability
Useful for a technical or executive evaluator assessing Precision measurement specifically; drawn from internal source evidence, not an independently public-published artifact.
Evidence
Evidence class: demonstrated-capability
Frequently asked
What is the evidence behind "Precision measurement: full census of 176 candidates, 2 defects fixed"?
PRECISION-POST-6510.md, which grades all 176 candidates individually against a stated standard.