RFR-2026-012: Cognitive-load and usability study

Research opportunity

Test whether explicit evidence, uncertainty, and provenance controls improve decisions without unacceptable cognitive load.

Background

The architecture assumes human/machine use, but usability and cognitive burden are not validated.

Origin documents and trace

• research/evaluations/FE-BOUNDARY-2026-07-28/RP-FE-BOUNDARY-2026-07-28.md, section: Recommended normative definition
• research/product-architecture/uncertainty-to-action-flow.md, section: uncertainty-to-action flow
• research/validation/validation-program.md, section: Claims Under Test

Specific assumption challenged: More explicit structure improves net usefulness for humans.

Supporting evidence: Traceability improvement is a claim under test; the boundary REP states net benefit is untested.

Unknowns

• Comprehension burden
• Expertise interaction
• Error-detection versus completion-time tradeoff

Dependencies

• RFR-2026-003

Suggested REP and methodology

Create REP-2026-012. Run task-based usability studies with novice/expert strata; measure comprehension, time, errors, confidence calibration, and abandonment; compare progressively disclosed variants.

Expected outputs

• Usability protocol
• Interaction logs
• Cognitive-load report

Success criteria

Benefits and harms are reported by expertise level with predefined minimally important differences.

Execution recommendation

• Recommended agent: human-computer-interaction agent
• Estimated effort: medium
• Expected knowledge gained: Tests a neglected human-use mechanism and informs profile simplification.
• Frontier Score: 184
• Score inputs: knowledge gain 4/5; impact 4/5; cross-project reuse 4/5; scientific importance 3/5; dependency cost 4/5; implementation difficulty 4/5.
• Score calculation: 4 × 4 × 4 × 3 − 4 − 4 = 184.
• Score status: structured expert judgment, not measured utility.