{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","source_assessment_id":"bounded_rivalry_governance__human_computer_interaction:P4:v0","cell_id":"bounded_rivalry_governance__human_computer_interaction","proposal_index":4,"qualification":"EMPIRICAL_PARTNER_CANDIDATE","criteria":{"specific_differentiated_claim":{"status":"YES","reason":"The proposal makes a falsifiable incremental claim that sealed, complementarity-based portfolio selection with equal resource ceilings yields more distinct, reproducible, repair-usable barriers per reviewer hour than first-valid allocation and a flat-fee panel."},"credible_problem_signal":{"status":"YES","reason":"Official accessibility guidance establishes consequential barriers and the limits of automated evaluation, while bounty research documents duplicate effort and first-reporter incentives; accessibility-specific strategic prevalence remains uncertain but the underlying signal is credible."},"identifiable_partner_or_adopter":{"status":"YES","reason":"Product owners and accessibility-program owners are concrete adopters, with security, privacy, and test-environment owners able to authorize access and independent accessibility evaluators or scoring partners able to conduct verification."},"partner_access_is_necessary":{"status":"YES","reason":"Resolving the causal claim requires an authorized prototype, scoped accounts, qualified testers, maintainer judgments of remediation utility, workflow timing, and safety observations that public research cannot supply."},"safe_authorized_first_step":{"status":"YES","reason":"The proposed first step is a nonproduction crossover mock using synthetic data, scoped accounts, guaranteed compensation, fictitious points, explicit prohibited actions, an urgent-disclosure path, stop conditions, and approval by the relevant product, accessibility, security, privacy, and environment owners."},"bounded_decisive_empirical_design":{"status":"YES","reason":"The two-round crossover compares the proposed mechanism with a flat-fee panel and independently rescored first-valid allocation, uses hidden seeded barriers and blinded scoring, and specifies measurable advancement and falsification outcomes."},"no_material_negative_gate":{"status":"YES","reason":"All substantive pipeline gates are YES. Cost scope is uncertain, but it does not block the bounded mock and is not the sole reason partner engagement is needed."},"not_merely_more_research":{"status":"YES","reason":"The next step specifies the environment, participants, comparators, crossover procedure, measurements, red-team exercises, advancement threshold, and falsifiers rather than requesting open-ended investigation."}},"uncertainty_types":["PROBLEM_PREVALENCE","ADOPTER_PULL","INCREMENTAL_EFFECT","WORKFLOW_FIT","DATA_ACCESS","COST_SCOPE"],"partner_profile":"An organization with a pre-release web-service prototype and an accessibility assurance program, represented by a product owner and accessibility lead who can recruit approximately 12 qualified compensated testers, obtain security/privacy/environment approvals, provide maintainers for remediation-utility ratings, and engage an independent scoring partner.","required_access":"Authorized access to an isolated resettable prototype with synthetic accounts and instrumented logs; qualified testers spanning assistive-technology and input-mode expertise; a hidden barrier seed register; blinded independent scorers; maintainer ratings and reviewer-time data; and approval to exercise submission, appeal, urgent-disclosure, stop, and adversarial-testing procedures.","bounded_empirical_test":"Run a preregistered two-round crossover mock on one isolated prototype. Randomize qualified testers between sealed competitive portfolio selection and a noncompetitive flat-fee panel, switch conditions on a reset fixture, and independently rescore the combined reports under first-valid allocation. Measure distinct reproduced barriers, seed recall, coverage, repair utility, reviewer time, scorer agreement, duplicates, fragmentation, burden, cap exceptions, appeals, disclosure latency, and safety events while red-teaming specified gaming behaviors.","success_condition":"The portfolio condition improves distinct, independently reproduced, repair-usable findings per reviewer hour by a preregistered material margin such as at least 20% over both comparators, while maintaining acceptable preregistered scorer reliability and showing no excess safety events, urgent-disclosure delay, tester burden, or disproportionate loss of manual or assistive-technology-intensive findings.","falsification_condition":"The claim is falsified if the flat-fee panel matches or exceeds the portfolio condition, first-valid allocation performs equivalently, complementarity scoring lacks preregistered reliability, caps suppress valuable manual or assistive-technology-intensive work, or competition increases burden, unsafe behavior, or urgent-report delay without a material coverage gain.","rationale":"This is a narrow empirical-partner case: adjacent practices support feasibility and the problem, but public evidence cannot resolve whether governed competition adds value over a well-run paid panel. A concrete authorized partner is necessary to supply the environment, participants, workflow data, and maintainer judgments for a safe comparator-based test that can either advance or reject the differentiated claim."}