evidence · EV-A11Y-2026-58E0
Research index
evidence
Documents grouped under evidence.
evidence · EV-AI-COMM-2026-3D6F
Greater abstractive summarization trades against factuality
evidence · EV-AI-COMM-2026-D120
Automatic factuality metrics do not reliably measure all factuality failures
evidence · EV-AICODING-2026-10A1
GitHub controlled HTTP-server task reported faster completion with Copilot
evidence · EV-AICODING-2026-3B8F
METR RCT found early-2025 AI slowed experienced maintainers
evidence · EV-AICODING-2026-6E21
DORA found positive individual perceptions alongside weaker delivery outcomes
evidence · EV-AICODING-2026-C940
Passing tests overstated agent merge readiness on realistic tasks
evidence · EV-AICODING-2026-F5D2
Later METR RCT became unreliable because AI abstention induced selection
evidence · EV-AREA-YYYY-0001
Replace with evidence title
evidence · EV-BASELINE-2026-2B6D
Replicable evaluation requires explicit intervention and comparator descriptions
evidence · EV-BASELINE-2026-A4F8
Crossover software experiments trade participant control for learning and carryover risk
evidence · EV-BOUNDARY-2026-01D3
Shannon formalizes reliable message transmission rather than meaning
evidence · EV-BOUNDARY-2026-274A
Human-centred design already governs interactive-system lifecycle activities
evidence · EV-BOUNDARY-2026-5E8B
Systems engineering already supplies integrative lifecycle methods
evidence · EV-BOUNDARY-2026-8C11
Information architecture covers findability and understandability
evidence · EV-BOUNDARY-2026-A792
Rhetoric already models action-bearing communication as situated
evidence · EV-COG-2026-27F1
Instructional support can reverse from helpful to harmful with expertise
evidence · EV-COG-2026-C5A4
Interruption cost depends on context timing source and priority
evidence · EV-CSCW-2026-32A9
Boundary-object management distinguishes shared from local artifacts
evidence · EV-CSCW-2026-9C40
Boundary artifacts require scope structure and management
evidence · EV-HANDOFF-2026-2A10
Unavailable coworkers block software-task information needs
evidence · EV-HANDOFF-2026-4F92
Documentation quality associates with stronger technical capability outcomes
evidence · EV-HANDOFF-2026-73D4
Architecture explanations depend on audience role experience and goal
evidence · EV-HANDOFF-2026-B6E8
Software onboarding outcomes are multidimensional and task-dependent
evidence · EV-IA-2026-E174
Findability and structural organization do not guarantee correct interpretation
evidence · EV-KM-2026-3E80
Transactive memory develops through team conditions and predicts performance
evidence · EV-KM-2026-91BC
Knowledge integration mediates transactive-memory performance effects
evidence · EV-MAINT-2026-0A52
Contextual resumption cues doubled interrupted-task success
evidence · EV-MAINT-2026-7B19
Information scent predicts maintenance navigation
evidence · EV-MEASURES-2026-11AF
Developer productivity cannot be represented by one activity or efficiency metric
evidence · EV-MEASURES-2026-5C7E
Comprehension studies separate time correctness and executable extension
evidence · EV-MEASURES-2026-9D30
Handoff effects must be assessed across recipient system and organizational outcomes
evidence · EV-REQ-2026-48A0
Requirements ambiguity tooling has weak empirical validation
evidence · EV-REQ-2026-D781
Industrial requirements transformation preserves decisions and deviations
evidence · EV-RHETORIC-2026-66B1
Commit-message quality is a situated collaborative communication problem
evidence · EV-SAFEGUARDS-2026-3A91
Privacy controls should follow an explicit map of data processing and individual risk
evidence · EV-SAFEGUARDS-2026-64C2
Human research requires consent risk-benefit assessment and equitable selection
evidence · EV-SAFEGUARDS-2026-D8E5
Uniform time limits can create accessibility barriers
evidence · EV-SAFETY-2026-19D5
Structured I-PASS handoffs reduced a composite adverse-event outcome
evidence · EV-SAFETY-2026-7EA2
Structured-handoff evidence is heterogeneous and often low certainty
evidence · EV-SYSTEMS-2026-22C8
Assurance arguments explain why evidence supports system claims
evidence · EV-SYSTEMS-2026-A730
Value of information measures decision improvement rather than information volume
evidence · EV-TECHCOMM-2026-14B2
Minimalist technical communication is task-oriented and user-centred
evidence · EV-TECHCOMM-2026-6FD1