{"schema_version":1,"research_id":"eoa_inverse_innovation_exp04_external_evaluation_20260802","source_assessment_id":"computability_boundary_mapping__ethnography_qualitative_methods:RETRIEVAL_FIRST:v0","cell_id":"computability_boundary_mapping__ethnography_qualitative_methods","search_queries":["site:fda.gov qualitative research saturation concept elicitation guidance saturation grid","site:nih.gov qualitative research saturation sampling guidance grounded theory","qualitative research saturation explicit uncertainty stopping no new codes future data critique Braun Clarke","qualitative data saturation software saturation monitoring product","PLOS ONE 2020 Guest Namey Chen simple method assess report thematic saturation full text","Hennink Kaiser Marconi 2017 Code Saturation Versus Meaning Saturation full text PDF","Nelson 2016 conceptual depth criteria saturation qualitative research PDF","Standards for Reporting Qualitative Research O'Brien 2014 full text SRQR","FDA patient focused drug development guidance 3 saturation qualitative data 2025 official","site:fda.gov \"saturation grid\" qualitative interviews FDA guidance","Beyond saturation Q-FORS qualitative framework operationalizing respondent sampling full text 2025","\"Defining, assessing, and reporting saturation\" 2026 PDF","qualitative saturation calculator software official Guest Namey Chen","site:help-nv.qsrinternational.com saturation qualitative coding NVivo","site:atlasti.com saturation qualitative research software saturation","site:dedoose.com saturation qualitative research software","HHS 45 CFR 46 secondary research identifiable private information exemption official","HHS OHRP coded private information not human subjects research guidance official archived data","BLS anthropologists archeologists median pay 2025 qualitative researcher","BLS software developers median pay 2025","site:bls.gov/ooh software developers quality assurance analysts testers 2024 median pay","site:bls.gov/ooh social science research assistants median pay 2024","qualitative research software pricing NVivo official 2026","ATLAS.ti pricing official 2026"],"sources":[{"source_id":"S1","title":"A simple method to assess and report thematic saturation in qualitative research","publisher":"PLOS ONE","url":"https://journals.plos.org/plosone/article/file?id=10.1371%2Fjournal.pone.0232076&type=printable","source_class":"PRIMARY_RESEARCH","publication_date":"2020-05-05","accessed_at":"2026-08-02","claims_supported":["Existing prospective saturation assessment uses base size, run length, and a new-information threshold.","Operationalizations vary and commonly produce binary saturation determinations.","Retrospective fixed-dataset measures necessarily approach 100 percent and do not prove future completeness.","The method was tested on three previously coded interview datasets."]},{"source_id":"S2","title":"Code Saturation Versus Meaning Saturation: How Many Interviews Are Enough?","publisher":"SAGE / Qualitative Health Research","url":"https://doi.org/10.1177/1049732316665344","source_class":"PRIMARY_RESEARCH","publication_date":"2016-09-25","accessed_at":"2026-08-02","claims_supported":["Code saturation can occur before meaning saturation.","Saturation claims are often inadequately justified or reported.","Samples that are too small can omit phenomena, while unnecessarily large samples waste resources and burden participants.","The study maintained chronological code-development records and distinguished descriptive from conceptual development."]},{"source_id":"S3","title":"Saturation in qualitative research: exploring its conceptualization and operationalization","publisher":"Springer Nature / Quality & Quantity","url":"https://link.springer.com/article/10.1007/s11135-017-0574-8","source_class":"AUTHORITATIVE_SECONDARY","publication_date":"2018-05-01","accessed_at":"2026-08-02","claims_supported":["Saturation has multiple incompatible conceptualizations across qualitative approaches.","Code recurrence can precede theoretical development or meaning saturation.","There is always potential for something new to emerge, making saturation an analyst judgment or matter of degree rather than literal completeness.","Published studies sometimes claim saturation and then collect additional cases to confirm it, exposing uncertainty in the terminal label."]},{"source_id":"S4","title":"Using conceptual depth criteria: addressing the challenge of reaching saturation in qualitative research","publisher":"Queen's University Belfast / SAGE","url":"https://pure.qub.ac.uk/en/publications/using-conceptual-depth-criteria-addressing-the-challenge-of-reach/","source_class":"PRIMARY_RESEARCH","publication_date":"2016-12-14","accessed_at":"2026-08-02","claims_supported":["Saturation has persistent definition and process problems.","Conceptual-depth criteria provide a structured evidence base for study-specific theoretical-sampling decisions.","Conceptual depth is an adequacy framework rather than proof that no future case could alter a theory."]},{"source_id":"S5","title":"Open-ended interview questions and saturation","publisher":"PLOS ONE","url":"https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0198606","source_class":"PRIMARY_RESEARCH","publication_date":"2018-06-20","accessed_at":"2026-08-02","claims_supported":["Empirical domain sizes have long tails, and most tested domains did not reach the study's saturation threshold within the original sample.","Bounded and unbounded elicitation domains behave differently.","Salience may be more useful than attempting exhaustive thematic coverage.","The study supplies empirical evidence that late or rare items can remain after apparent early coverage."]},{"source_id":"S6","title":"Patient-Focused Drug Development: Methods to Identify What Is Important to Patients","publisher":"U.S. Food and Drug Administration","url":"https://www.fda.gov/media/131230/download","source_class":"OFFICIAL_GUIDANCE","publication_date":"2022-02-01","accessed_at":"2026-08-02","claims_supported":["FDA recognizes saturation as a possible stopping principle but states that there are no set criteria or methodology for evaluating it.","FDA describes consecutive interview-set comparisons and saturation grids as one approach.","FDA recommends chronological concept tracking, quality checks, audit trails, and attention to target-population coverage.","FDA is an identifiable authorizer of qualitative evidence used in patient-focused drug development."]},{"source_id":"S7","title":"Evaluating Respiratory Symptoms (E-RS) in COPD: Clinical Outcome Assessment Staff Review","publisher":"U.S. Food and Drug Administration","url":"https://www.fda.gov/media/109854/download","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"2016-01-01","accessed_at":"2026-08-02","claims_supported":["FDA staff reviewed submitted saturation evidence, transcripts, ATLAS.ti summaries, and a chronological saturation grid.","The review accepted a conclusion based on three focus groups after no new symptoms emerged.","Regulatory review demonstrates consequential use of bounded no-new-concept evidence and identifies a concrete authorizer and workflow."]},{"source_id":"S8","title":"Standards for reporting qualitative research: a synthesis of recommendations (SRQR)","publisher":"EQUATOR Network","url":"https://www.equator-network.org/reporting-guidelines/srqr/","source_class":"STANDARD","publication_date":"2014-09-01","accessed_at":"2026-08-02","claims_supported":["SRQR is a 21-item reporting standard applying to whole qualitative research reports.","The standard creates an established reporting venue for explicit sampling rationale, methods, and limitations.","Journal editors, reviewers, and research institutions are identifiable diffusion channels for auditable stopping records."]},{"source_id":"S9","title":"Application of ATLAS.ti in qualitative data analysis","publisher":"ATLAS.ti","url":"https://atlasti.com/research-hub/the-application-of-atlas-ti-in-different-qualitative-data-analysis-strategies","source_class":"COMMERCIAL_FIRST_PARTY","publication_date":"n.d.","accessed_at":"2026-08-02","claims_supported":["A major qualitative-analysis vendor supports flexible coding workflows, iterative evaluation, and memo-based process documentation.","Its guidance describes coding until saturation and leaves methodological accountability with the researcher.","The reviewed first-party material does not expose a mandatory three-state terminal contract."]},{"source_id":"S10","title":"Automatic coding using existing coding patterns","publisher":"Lumivero / NVivo","url":"https://help-nv.qsrinternational.com/20/win/Content/coding/automatic-coding-existing-patterns.htm?Highlight=autocode","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"n.d.","accessed_at":"2026-08-02","claims_supported":["NVivo supports iterative coding of successive datasets and preserves provenance for automatically coded references.","NVivo warns that pattern-based coding is approximate and unsuitable for interpretive concepts, attitudes, tones, or emotions.","Human coding and review remain necessary for theory-altering novelty judgments."]},{"source_id":"S11","title":"Coded Private Information or Specimens Use in Research, Guidance","publisher":"U.S. Department of Health and Human Services, Office for Human Research Protections","url":"https://www.hhs.gov/ohrp/regulations-and-policy/guidance/research-involving-coded-private-information/index.html","source_class":"OFFICIAL_GUIDANCE","publication_date":"2008-10-16","accessed_at":"2026-08-02","claims_supported":["Some secondary research using coded private information is not human-subjects research when investigators cannot readily identify individuals and the data were not collected for the proposed project.","Identifiable archived data can require exemption analysis or IRB review.","A shadow study should use deidentified or properly governed records and obtain an institutional determination."]},{"source_id":"S12","title":"45 CFR 46","publisher":"U.S. Department of Health and Human Services, Office for Human Research Protections","url":"https://www.hhs.gov/ohrp/regulations-and-policy/regulations/45-cfr-46/index.html?locale=us","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"2024-10-01","accessed_at":"2026-08-02","claims_supported":["The Common Rule establishes protections and institutional-review requirements for covered human-subjects research.","Additional protections apply to specified populations.","Live deployment does not remove the study lead's or institution's legal and ethical authority."]},{"source_id":"S13","title":"Software Developers, Quality Assurance Analysts, and Testers","publisher":"U.S. Bureau of Labor Statistics","url":"https://www.bls.gov/ooh/computer-and-information-technology/software-developers.htm","source_class":"OFFICIAL_ORGANIZATION_DATA","publication_date":"2025-08-28","accessed_at":"2026-08-02","claims_supported":["The May 2024 median annual wage was $133,080 for software developers and $102,610 for software quality-assurance analysts and testers.","Software development normally includes requirements analysis, testing, security consideration, maintenance, and stakeholder feedback.","These wage levels support the labor assumptions used in the broad cost bands."]},{"source_id":"S14","title":"NVivo Software","publisher":"York University Information Technology","url":"https://www.yorku.ca/uit/nvivo/","source_class":"OFFICIAL_ORGANIZATION_DATA","publication_date":"2026-01-01","accessed_at":"2026-08-02","claims_supported":["A 2026 institutional NVivo license costs $250 for faculty and staff and $180 for students at this institution.","Existing institutional qualitative-analysis software can keep a small shadow study's direct tooling cost low.","The license covers organization, management, coding, and analysis of qualitative data."]},{"source_id":"S15","title":"Beyond saturation: A qualitative framework for operationalizing respondent sampling (Q-FORS) for data adequacy","publisher":"Elsevier / Social Science & Medicine","url":"https://pubmed.ncbi.nlm.nih.gov/41391293/","source_class":"AUTHORITATIVE_SECONDARY","publication_date":"2025-11-01","accessed_at":"2026-08-02","claims_supported":["Saturation has been problematically generalized beyond methodological contexts where it is appropriate.","Q-FORS supplies a transparent data-adequacy framework spanning proposal, collection, and reporting stages.","The framework is close prior art for structured sufficiency decisions but does not, in the reviewed abstract, mandate finite-frame exhaustion and explicit unknown states."]},{"source_id":"S16","title":"Q-Sat AI: Machine Learning-Based Decision Support for Data Saturation in Qualitative Studies","publisher":"arXiv","url":"https://arxiv.org/abs/2511.01935","source_class":"PRIMARY_RESEARCH","publication_date":"2025-11-02","accessed_at":"2026-08-02","claims_supported":["A recent preprint proposes machine-learning decision support for qualitative saturation and sample-size justification.","The proposed users include qualitative researchers, journal reviewers, and thesis advisors.","It is adjacent computational prior art but predicts or standardizes saturation rather than enforcing the candidate's finite-frame/unknown distinction."]}],"problem_evidence":{"support":"STRONG","rationale":"Multiple independent empirical and methodological sources document vague, inconsistent, or weakly justified saturation claims, distinguish early code recurrence from later meaning or theoretical development, and show long-tailed discovery. FDA guidance confirms that saturation affects consequential regulatory evidence while acknowledging the absence of set evaluation criteria. The narrower allegation that existing dashboards routinely cause premature stopping is not directly quantified and remains an evidence gap.","source_ids":["S1","S2","S3","S4","S5","S6","S15"]},"stakeholder_evidence":{"support":"MODERATE","rationale":"FDA is an identifiable authorizer that requests chronological saturation evidence in patient-focused work, while study leads, reviewers, EQUATOR-aligned journals, and QDA vendors have relevant decision or implementation authority. Sources express demand for transparent, methodical stopping rationales, but no reviewed adopter explicitly requests the proposed three-state terminology or commits to deploying it.","source_ids":["S2","S6","S7","S8","S9","S15","S16"]},"prior_art":{"proximity":"SUBSTANTIAL_COLLISION","closest_analogues":[{"name":"Guest, Namey, and Chen prospective saturation assessment","similarity":"Uses a declared base, run length, novelty threshold, chronological coding evidence, and prospective stopping calculation, covering most of the candidate's bounded-evidence workflow.","remaining_difference":"It reports saturation levels rather than mandating UNKNOWN_AFTER_BOUND, and it does not reserve exact exhaustion for a verified finite candidate frame.","source_ids":["S1"]},{"name":"FDA saturation-grid workflow","similarity":"Uses consecutive groups, chronological concept emergence, population-coverage review, audit trails, and consequential stopping evidence in an authorized regulatory workflow.","remaining_difference":"FDA still uses a saturation label and does not require a three-state interface or a finite-frame exhaustion certificate.","source_ids":["S6","S7"]},{"name":"Conceptual depth and Q-FORS data-adequacy frameworks","similarity":"Replace literal completeness with transparent, study-relative sufficiency reasoning and structured evidence for sampling decisions.","remaining_difference":"They remain researcher-governed adequacy frameworks rather than machine-enforced output semantics with UNKNOWN and finite-frame-only exhaustion.","source_ids":["S4","S15"]},{"name":"ATLAS.ti/NVivo iterative qualitative-analysis workflows","similarity":"Already support chronological coding, memos, provenance, iterative datasets, and human review in the target software environment.","remaining_difference":"Reviewed documentation does not enforce the candidate's state machine, and NVivo explicitly limits automation of interpretive judgments.","source_ids":["S9","S10"]},{"name":"Q-Sat AI saturation decision support","similarity":"Directly proposes computational support for saturation and sample-size decisions for researchers and reviewers.","remaining_difference":"It models saturation estimates rather than refusing universal completeness claims through finite-exhaustion and explicit-unknown semantics.","source_ids":["S16"]}],"distinctive_claim_remaining":"For each study-round stopping decision, an enforceable three-state protocol that issues FINITE_FRAME_EXHAUSTED only after complete accounting of an enforceably finite eligible frame, otherwise reports a bound-ending no-novelty run as UNKNOWN_AFTER_BOUND, will reduce premature terminal saturation decisions relative to an ordinary bounded saturation workflow without causing routine human relabeling or bypass.","confidence":"HIGH"},"implementation_evidence":{"support":"MODERATE","rationale":"The state machine, eligibility ledger, chronology, bounds, and versioned audit record are conventional software constructs and fit existing QDA workflows. The technically hard part is not computation but semantic validity: determining whether the eligibility frame is substantively complete and whether a case is theory-altering requires qualified human judgment. Archived data can be used safely if deidentified or institutionally authorized. Production use would require access controls, immutable provenance, local IRB/privacy review where applicable, and a strict rule that software cannot stop recruitment autonomously.","source_ids":["S6","S7","S9","S10","S11","S12","S13"]},"scores":{"meaningful_impact":{"score":4,"rationale":"Avoiding overstated theoretical completeness could protect research validity, represented populations, participant burden, and downstream regulatory or policy decisions; actual prevalence and realized impact are unmeasured.","source_ids":["S2","S3","S5","S6","S7"]},"stakeholder_pull":{"score":3,"rationale":"Regulators, reviewers, and researchers clearly need transparent stopping evidence, but there is no direct request or commitment for this exact state contract.","source_ids":["S6","S7","S8","S15"]},"incremental_advantage":{"score":3,"rationale":"Mandatory UNKNOWN and finite-only exhaustion are clearer than existing saturation grids, but much of the audit trail and sufficiency reasoning already exists, and comparative benefit has not been tested.","source_ids":["S1","S4","S6","S15"]},"distinctiveness_plausibility":{"score":3,"rationale":"No reviewed source combined both enforcement conditions, but the contribution is principally output semantics and governance layered over substantial prior art; worldwide novelty is unmeasured.","source_ids":["S1","S4","S6","S9","S15","S16"]},"technical_implementability":{"score":4,"rationale":"A deterministic router, ledger, and versioned record are straightforward to implement; reliable human classification of novelty and frame completeness is the main limitation.","source_ids":["S9","S10","S13"]},"adoption_authority_feasibility":{"score":3,"rationale":"Study leads and software maintainers can authorize shadow use, and FDA-regulated teams have an established evidence workflow. Live recruitment decisions remain with investigators and institutional oversight, and regulator acceptance of the new labels is unknown.","source_ids":["S6","S7","S11","S12"]},"evidence_readiness":{"score":3,"rationale":"The comparator, outcomes, traces, and falsifiers are specifiable now, but no prototype, retrospective benchmark, frame-classification reliability result, or adopter commitment has been produced.","source_ids":["S1","S2","S5","S9"]},"safety_net_benefit":{"score":4,"rationale":"An explicit UNKNOWN state is a useful non-deceptive fallback that prevents software from turning a timeout or bounded plateau into a universal negative; human users may still bypass or misread it.","source_ids":["S3","S6","S10"]},"scalability":{"score":3,"rationale":"The record structure is portable across QDA systems, but different qualitative traditions reject or interpret saturation differently, so broad scaling requires method-specific configuration and governance.","source_ids":["S3","S8","S9","S15"]}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"10K_TO_50K","scope":"Preregister and execute the 20-trace paired shadow evaluation, including trace preparation, two qualified reviewers, adjudication, analysis, and a short methods report.","confidence":"MODERATE","assumptions":["Use archived deidentified or synthetic traces; no participant recruitment.","Approximately 150-300 combined hours of qualitative-methodologist, analyst, and coordination labor.","Existing institutional QDA licenses and ordinary secure storage are available.","No production integration is included."],"source_ids":["S11","S13","S14"]},"initial_deployment_startup":{"band_2026_usd":"50K_TO_250K","scope":"Build and validate an MVP plugin or companion application with the three-state router, eligibility ledger, audit log, role permissions, exports, and tests for one QDA workflow.","confidence":"MODERATE","assumptions":["Roughly 0.5-1.5 software-engineer years plus fractional methodologist, UX, QA, and security effort.","The MVP imports structured exports rather than modifying a proprietary QDA core.","No autonomous coding or recruitment control is implemented.","Existing authentication and hosting services are reused."],"source_ids":["S9","S10","S13"]},"operational_launch":{"band_2026_usd":"250K_TO_1M","scope":"Production launch across several research teams or institutions, including vendor integration, security/privacy assessment, validation, training, documentation, support setup, and institutional governance review.","confidence":"LOW","assumptions":["Multiple QDA formats and institutional security requirements are supported.","Archived and live-project data require access controls, provenance, retention rules, and local legal/IRB review.","Qualified methodologists validate frame and novelty workflows.","The system remains advisory and cannot autonomously stop recruitment."],"source_ids":["S6","S10","S11","S12","S13"]},"annual_recurring":{"band_2026_usd":"50K_TO_250K","scope":"Maintenance, security updates, compatibility testing, user support, method-governance review, audit retention, and periodic training for a modest institutional deployment.","confidence":"LOW","assumptions":["Approximately 0.4-1.0 combined engineering/support FTE plus fractional methodologist and privacy oversight.","Cloud volume is modest because source transcripts need not be centrally duplicated.","Vendor and institutional license costs are additional where not already covered.","No large-scale recruiting or manual coding service is included."],"source_ids":["S13","S14"]}},"verified_pipeline_gates":{"externally_supported_problem":{"status":"YES","reason":"Independent research and official guidance document ambiguity, overstatement, inconsistent operationalization, and consequential use of bounded saturation evidence.","source_ids":["S1","S2","S3","S5","S6"]},"externally_credible_adopter_or_authorizer":{"status":"YES","reason":"FDA is an identifiable authorizer of qualitative evidence, study leads control sampling decisions, and established QDA vendors and reporting-guideline users are credible implementation channels. Direct commitment to this intervention remains absent.","source_ids":["S6","S7","S8","S9"]},"distinct_testable_incremental_claim":{"status":"YES","reason":"The residual claim compares mandatory three-state semantics against ordinary bounded saturation on premature terminal decisions, invalid exhaustion certificates, UNKNOWN relabeling, and frame-classification agreement.","source_ids":["S1","S4","S6"]},"bounded_next_evidence_step":{"status":"YES","reason":"A preregistered 20-trace paired shadow evaluation has a fixed sample, comparator, metrics, stop condition, and explicit falsifiers, with no effect on live recruitment.","source_ids":["S1","S2","S5"]},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The shadow study can be performed without changing participant contact or recruitment, provided records are deidentified or an institutional determination authorizes their use. Live deployment would require separate study-lead and institutional authority.","source_ids":["S11","S12"]},"credible_cost_scope_and_range":{"status":"YES","reason":"Each band is tied to a defined scope and staffing assumptions, anchored by official software labor data and a current institutional QDA license price; launch and recurring estimates remain low-confidence because integration scope is unknown.","source_ids":["S13","S14"]}},"next_evidence_step":"Preregister a paired shadow test using 20 fixed trajectories: 10 with a declared finite candidate ledger and 10 with open-ended theoretical sampling. Include a delayed admissible theory-altering or negative case after an apparent no-new-code plateau wherever logically possible. Two blinded qualified reviewers independently apply (A) the Guest-style base-size/run-length/new-information-threshold workflow and (B) the three-state protocol to identical round histories, then make a separate human continue/stop judgment. Measure premature terminal saturation judgments before the delayed case, invalid FINITE_FRAME_EXHAUSTED certificates, frame-classification agreement, UNKNOWN frequency, human relabeling/bypass, extra rounds requested, and decision time. Stop at 20 traces. Advance only if the protocol produces zero invalid exhaustion certificates, reduces premature terminal judgments by at least 50 percent relative to baseline, attains frame-classification agreement of kappa at least 0.70, and has no more than 25 percent of UNKNOWN outputs relabeled as saturation. Falsify or redesign if any invalid certificate occurs, premature decisions are not reduced, agreement is below 0.70, or more than 25 percent of UNKNOWN outputs are bypassed or relabeled.","blocking_evidence":["No direct study quantifies how often current qualitative-research software labels cause premature live recruitment stopping or are interpreted as universal completeness.","No matched-trace result shows that the proposed semantics improve authorized human decisions rather than merely changing labels.","No evidence establishes reliable reviewer classification of finite versus open-ended frames or of theory-altering novelty.","A formally finite ledger can omit substantively eligible people; the intervention has no demonstrated safeguard against a misleadingly narrow frame definition.","No FDA, journal, university, or QDA vendor has committed to accept or deploy the proposed labels.","Patentability, freedom to operate, market size, worldwide novelty, and realized impact were not measured."],"research_disposition":"PARTNERED_RESEARCH_PROGRAM","world_novelty_boundary":"Bounded web review establishes that saturation criteria, prospective run rules, saturation grids, conceptual-depth/data-adequacy frameworks, chronological audit trails, and computational saturation support are prior art. The only plausible residual distinction is the jointly enforced three-state output alphabet plus finite-frame-only exhaustion certificate. Search failure does not establish worldwide novelty; patents, non-English sources, proprietary product behavior, freedom to operate, market size, and realized impact remain unmeasured.","arm":"RETRIEVAL_FIRST","candidate_version":0,"controller_recommendation":{"action":"STOP_EMPIRICAL_RESEARCH_NEEDED","repairable":true,"material_progress_observed":false,"progress_targets":["Complete the preregistered 20-trace paired shadow evaluation with the stated thresholds and falsifiers.","Demonstrate zero invalid exhaustion certificates and at least 50 percent fewer premature terminal judgments than the bounded-saturation comparator.","Establish inter-reviewer finite/open-frame classification agreement of kappa at least 0.70 and qualitatively analyze every disagreement.","Show that no more than 25 percent of UNKNOWN outputs are relabeled or bypassed by authorized reviewers.","Obtain a written adoption or evaluation commitment from at least one qualitative-methods center, regulated patient-research team, journal methods group, or QDA vendor.","Secure an institutional privacy/IRB determination for any archived identifiable traces before analysis."],"reason":"Bounded web research has resolved the visible problem, adopter classes, substantial prior-art collision, technical plausibility, authority constraints, and a testable residual claim. The decisive remaining evidence is behavioral and empirical: qualified reviewers must use the competing workflows on matched traces, including delayed negative cases. That evidence cannot be supplied by further web search alone."}}