Confidence and reliability flags

Every AS measurement carries a confidence level computed from data completeness, collection reliability, and category fit. Datapoint scores within the AS measurement inherit this confidence.

A high-confidence AS measurement reports datapoint scores that the practitioner can act on. A low-confidence measurement reports datapoint scores that should be treated as preliminary; remedy is to expand data acquisition rather than to act on the preliminary findings.

Confidence affects how the practitioner reads each datapoint:

Confidence	Datapoint score interpretation
High	Score reflects reality; act on it
Medium	Score is directionally accurate; trends across cycles are reliable, single-cycle absolutes less so
Low	Score is preliminary; expand data acquisition before acting
Insufficient	Score is uninformative; do not act on it

The Reference’s nine-section template is written assuming high-confidence measurement. When confidence is lower, the practitioner reads each section with the corresponding caveat applied.

---