Confidence and Uncertainty
Last updated 10/7/2026
Confidence in this course describes the quality of the evidence record. It is not a probability that an outside-world event will occur.
The rubric is an explicit UAYA course convention. No received source cited in this act presents these high/medium/low tiers as a validated measure.
Evidence-quality rubric
| Tier | Required conditions | Wording consequence |
|---|---|---|
| High | target settled; exact inputs; every material rule cited or completely derived in the reader-facing worksheet; no unresolved conflict that could change the result; relevance trace complete | state the provisional course direction and its limit plainly |
| Medium | inputs and target settled; one bounded ambiguity remains, but it does not change the current rule family; alternatives are visible | use provisional wording and name the alternative |
| Low | a material rule conflict, role ambiguity, or incomplete relevance path could change the result | do not state a clean direction; repair the record or abstain |
“High” does not mean certain. “Low” is not a license to guess more softly.
Apply the rubric
Example type: three versions of one invented record.
- High evidence quality: exact Ganzhi; one target; reproducible states; each applied rule or course convention named; all active factors traced; no material disagreement.
- Medium evidence quality: same record, but a visible harm relation has no independent predictive weight under the cited chapter. Record it as visible background, state that it does not change the comparison, and lower the tier for the bounded secondary ambiguity.
- Low evidence quality: two equally eligible obstacle candidates remain after the selection lesson, and their recorded states would produce different comparison rows. Record
low; abstain pending role selection.
The confidence note should answer four questions:
- Which inputs were exact?
- Which rule set was applied?
- Which material alternatives were rejected, and why?
- What would change the current result?
No pseudo-precision
Do not convert the tier into a percentage or rating. No numerical confidence measure is supplied by this rubric.
Trap and stopping condition
- Do not award high confidence because the prose sounds coherent.
- Do not hide a disagreement between a cited source and a declared course convention in a footnote.
- Do not label an unresolved material conflict “medium” merely to keep a direction.
- Stop and abstain when repairing the missing evidence could change the rule family.
Source note
This rubric is an explicit UAYA teaching convention. It uses the evidence fields built in Acts 4–7 and is not presented as a classical doctrine, a probability, or an efficacy measure.
Next, audit every retained active factor: Movement Has Cause.