research-document HYR-VE-PE-001
Perceptual Envelope Hypothesis Registry
Perceptual Envelope Hypothesis Registry
| ID | Hypothesis | Initial confidence | Primary falsification experiment |
|---|---|---|---|
| HY-VE-PE-001 | A measured perceptual bottleneck predicts benefit from a visual intervention better than disability diagnosis alone. | Medium | EX-VE-TYP-002 |
| HY-VE-PE-002 | A meaningful subset of users has a stable-enough individual readability envelope that can be recovered from repeated measurements. | Medium | EX-VE-TYP-001 |
| HY-VE-PE-003 | The best visual configuration differs by task class, especially prose versus low-context identifiers. | Medium-high | EX-VE-TYP-001 |
| HY-VE-PE-004 | Semantic channel survivability predicts state-identification robustness under color loss, CVD, and forced colors. | Medium-high | EX-VE-COL-006 |
| HY-VE-PE-005 | A calibrated adaptive visual profile outperforms fixed named accessibility modes for heterogeneous users on at least one task-performance dimension without reducing comprehension. | Medium | EX-VE-X-002 |
| HY-VE-PE-006 | User preference is an imperfect proxy for best measured visual performance. | High | EX-VE-TYP-001, EX-VE-X-002 |
| HY-VE-PE-007 | Minimax palette optimization reduces worst-case categorical confusion across simulated observer models, but redundant semantics remain necessary for affected-user robustness. | Medium | EX-VE-COL-006 |
| HY-VE-PE-008 | Short-task equivalence can conceal compensation cost that appears as fatigue, regressions, slower recovery, or reduced persistence over sustained use. | Medium | Future sustained-reading experiment |
Rules
- A disability diagnosis may be used for recruitment, stratification, and subgroup analysis.
- A diagnosis must not be treated as proof of the active perceptual bottleneck.
- Preference and performance outcomes must be recorded separately.
- Simulation results must be labeled as simulation.
- Null and adverse results must remain in the registry.
- A hypothesis is not promoted to a principle until replicated or supported by converging evidence.