Binary success/failure presentation hides intermediate uncertainty

What it is

Binary uncertainty collapse reduces continuous or multi-source evidence to success/failure, detected/not detected, or safe/unsafe. Cases just across a threshold appear categorically different, while a fragile pass looks identical to overwhelming evidence.

Why it happens

Binary output supports quick action but removes margin, competing candidates, and missing data. People cannot identify borderline cases and rely equally on all “successes.” Changing thresholds can also flip identical input without visible cause.

Studying it

Binary, three-level, interval, and candidate-set displays can be compared near thresholds on verification, delegation, correction, and time. Equal outcome accuracy with varied evidence strength is needed, separating ambiguity, missing data, and model disagreement. Added decision burden should be reported.

Where it stops holding

Emergency operations may need a final binary action, but intermediate state should still change speed, resensing, and redundancy. Too many levels become a learned code. A middle state matters only when connected to action.

Applying it

• Preserve near-threshold, close-candidate, and missing-input states rather than merging them into failure.
• Bind each to verification, added input, degradation, or stop, with access to evidence.
• Test borderline cases for confident error, unnecessary confirmation, and decision time to find the smallest effective set.

Related

• Same group: X3.08.1 Perceptual and decision confidence must be presented in human-interpretable form · X3.08.3 Expressing uncertainty should not undermine overall system credibility · X3.08.4 Low-confidence situations should request confirmation rather than execute silently at a threshold
• Adjacent: X3.01 Visible current state · X3.05 Task progress and remaining work
• Search terms: binary uncertainty · threshold effect · intermediate state