{"schema_version":1,"assessment_id":"eoa_inverse_innovation_exp03_opportunity320_20260801","source_experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"invariant_mode_decomposition_design__ethnography_qualitative_methods","archetype_slug":"invariant_mode_decomposition_design","domain_slug":"ethnography_qualitative_methods","title":"Coupled-transition mode guidance for longitudinal qualitative analysis","opportunity_summary":"Evaluate whether a bounded transition-mode analysis can reveal persistent, escalating, or reversing configurations in ordered ethnographic episodes that thematic coding and sequence-aware manual comparison miss, while preserving residuals for negative-case analysis. Usefulness depends on defensible state construction, reproducibility under alternative coding and segmentation, and demonstrated incremental analytic value.","adopter_authorizer":"The adopting unit is a research team analyzing an already-authorized longitudinal corpus; the principal investigator, data steward, and applicable ethics oversight authorize use within consent and governance constraints.","scores":{"meaningful_impact":{"score":4,"rationale":"If coupled configurations are missed, the proposal identifies consequential effects on explanation, negative-case recognition, and theoretical sampling. Impact could therefore be substantial within affected studies, although the prevalence and frequency of this problem are unsupported."},"stakeholder_pull":{"score":2,"rationale":"Ethnographers and qualitative analysts are identifiable potential users, but the sealed candidate contains no interviews, requests, adoption signals, budget commitments, or evidence that they prefer a modal workflow over established qualitative practices."},"incremental_advantage":{"score":3,"rationale":"The proposal specifies potential gains over both individual-code review and sequence-aware manual comparison in held-out reconstruction, consequential negative-case discovery, and documented analytic revision. None of those gains has yet been demonstrated."},"distinctiveness_plausibility":{"score":3,"rationale":"The combination of transition-operator decomposition, held-out testing, residual routing, and qualitative interpretive safeguards is internally differentiated from the stated workflows, but prior art is explicitly unsearched and world-level distinctiveness cannot be inferred."},"technical_implementability":{"score":3,"rationale":"A retrospective read-only analysis on an existing ordered corpus is technically plausible, but defensible episode states, adequate sample size, operator stability, non-normal behavior, and sensitivity to coding and segmentation remain material implementation risks."},"adoption_authority_feasibility":{"score":4,"rationale":"The research team can control analytic use for an already-authorized corpus, with data-steward and ethics constraints clearly acknowledged. Feasibility falls short of the maximum because consent scope, governance approval, and downstream characterization risks must be checked for each corpus."},"evidence_readiness":{"score":3,"rationale":"The candidate supplies explicit baselines, a nearest rival, held-out evaluation, negative tests, halt criteria, and rollback. Readiness remains mixed because outcome rubrics, minimum data adequacy, independent-coding procedures, and stability thresholds are not operationalized."},"safety_net_benefit":{"score":4,"rationale":"The first step is read-only and reversible, modes are bounded as descriptive aids, residuals are directed to negative-case analysis, and participant ranking and field manipulation are excluded. Sparse configurations and reifying interpretations still create disclosure and characterization risks."},"scalability":{"score":2,"rationale":"Each corpus may require bespoke episode definition, coding-state construction, interpretive validation, sensitivity analysis, governance review, and analyst judgment. The local-stability assumption also limits straightforward transfer across sites, periods, or coding schemes."}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"10K_TO_50K","scope":"One retrospective, read-only pilot on an already-authorized corpus, including data preparation, earlier-episode fitting, held-out comparison against the baseline and nearest rival, blinded analyst review, sensitivity checks, and a decision memo.","confidence":"LOW","assumptions":["The corpus and primary coding record already exist and are accessible.","No new participant recruitment or data collection is required.","A small interdisciplinary team can complete the analysis with existing statistical software or modest custom code.","Independent recoding is limited to a validation subset rather than the full corpus."]},"initial_deployment_startup":{"band_2026_usd":"50K_TO_250K","scope":"Prepare a repeatable workflow for one research program, including documented state construction, analysis code, access controls, disclosure review, analyst training, audit records, and validation protocols.","confidence":"LOW","assumptions":["Deployment remains analyst-facing and does not automate conclusions or participant-facing decisions.","The institution already has secure qualitative-data infrastructure.","One or a few related corpora are in scope.","Substantial custom software product development is not required."]},"operational_launch":{"band_2026_usd":"50K_TO_250K","scope":"Launch governed use across an initial set of studies, including corpus-specific configuration, comparative analyses, ethics and data-steward coordination, analyst support, stability testing, and launch evaluation.","confidence":"LOW","assumptions":["Several studies participate but remain within one institution or consortium.","Existing consent permits the proposed secondary analytic processing.","Human analysts review all outputs and retain interpretive authority.","No new field collection is initiated solely from modal output during launch."]},"annual_recurring":{"band_2026_usd":"50K_TO_250K","scope":"Annual operation for a small research program, including analyst labor, corpus onboarding, governance review, secure computing, software maintenance, monitoring for instability or drift, and periodic evaluation against qualitative rivals.","confidence":"LOW","assumptions":["Usage is limited to a modest portfolio of longitudinal corpora.","Recurring expense is dominated by expert labor rather than compute.","Each new corpus requires contextual validation and disclosure review.","No licensing, compliance, or data-access obligation outside the sealed candidate is assumed."]}},"research_burden":"HIGH","earliest_credible_horizon":"3_TO_12_MONTHS","pipeline_gates":{"recognizable_externally_supportable_problem":{"status":"YES","reason":"The candidate identifies a recognizable analytic failure: individual-code summaries can miss reproducible coupled temporal configurations, with observable consequences for interpretation and sampling and a stated falsifier. Its prevalence remains unmeasured but is not required to recognize the bounded problem."},"identifiable_adopter_or_authorizer":{"status":"YES","reason":"The research team is the identifiable analytic adopter and has authority over use within participant consent, data-governance, data-steward, and ethics-review constraints."},"distinct_testable_incremental_claim":{"status":"YES","reason":"Modal guidance is explicitly compared with ordinary thematic analysis and sequence-aware manual case comparison on held-out reconstruction, negative-case discovery, and documented analytic revision."},"bounded_next_evidence_step":{"status":"YES","reason":"The candidate authorizes a retrospective, read-only, single-corpus pilot using earlier episodes for fitting, a held-out block for testing, blinded method comparison, stability checks, and explicit halt criteria."},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The bounded first step excludes participant ranking, new collection, field manipulation, automated conclusions, and publication of modes as essences; it also preserves rollback. Corpus-specific consent and disclosure checks remain prerequisites rather than an unresolved inherent stop."},"implementation_cost_scope_and_range":{"status":"UNCERTAIN","reason":"A bounded pilot scope is present, but the sealed candidate provides no corpus size, coding condition, staffing, software requirement, institutional overhead, or rollout scale from which implementation resources could be validated; the reported bands therefore rest on explicit assumptions."}},"blocking_evidence":["Reproducible coupled temporal structure beyond individual-code summaries and sequence-aware manual comparison on held-out and independently coded material.","Stability of selected modes under reasonable changes to coding, episode segmentation, observation density, time window, and model specification.","Incremental improvement in at least one prespecified analytic outcome without degradation in consequential negative-case or residual treatment.","Evidence that sparse configurations can be handled without unacceptable disclosure or harmful characterization risk.","Prior-art research sufficient to determine whether the proposed transfer is meaningfully distinct from existing qualitative, mixed-methods, spectral, dynamic-state, or transition-operator approaches.","Direct evidence that qualitative research teams consider the problem important enough to adopt and govern the added workflow."] ,"next_evidence_step":"Conduct one preregistered, retrospective, read-only pilot on an already-authorized longitudinal corpus. Fit only on earlier episodes and evaluate a held-out block; use an independently coded validation subset and blind reviewers to method labels while comparing ordinary thematic analysis, sequence-aware manual comparison, and modal guidance. Prespecify reconstruction, consequential negative-case discovery, documented analytic revision, residual concentration, and stability criteria. Falsify progression if coupled structure is not reproducible beyond both comparators, modal guidance adds no analytic value, or conclusions change qualitatively under reasonable coding or segmentation choices.","research_questions":["Can analysts define ordered episodes and meaning-bearing states without removing contextual or reflexive information needed for interpretation?","Does held-out and independently coded material contain reproducible coupled transitions beyond both stated comparison workflows?","Are selected configurations stable across reasonable coding, segmentation, observation-density, time-window, and operator choices?","Does modal guidance improve consequential negative-case discovery or documented analytic revision without suppressing rare accounts or redirecting attention away from participant priorities?","What consent, governance, and disclosure conditions are necessary when sparse configurations may indirectly identify participants?","What relevant prior art exists, and does this proposal offer a distinct incremental contribution after that boundary is established?","Do prospective adopter teams perceive sufficient unmet need and workflow fit to justify the added analytical and governance burden?"],"recommendation":"PARTNERED_RESEARCH","uncertainty_constraints":["Problem prevalence, stakeholder demand, market size, realized impact, and world-level distinctiveness are unmeasured.","Prior-art status is explicitly UNSEARCHED, so no novelty claim is supportable.","The proposal is a hypothesis with no reported pilot result or demonstrated advantage over either comparator.","Feasibility depends on an adequately sized, ordered, already-authorized corpus and a defensible coded-state representation.","Cost bands are resource-equivalent planning ranges based on stated assumptions, not observed prices or point estimates.","Local transition stability may fail because ethnographic meaning can be contextual, reflexive, discontinuous, and influenced by prompts or observation schedules.","Adoption authority is limited to analytic use and does not authorize new collection, participant labeling, field intervention, or causal and ontological claims."],"closed_book_prior_art_boundary":"No external search or source was used. The sealed packet states that prior art is unsearched and world novelty is unmeasured; therefore this assessment makes no claim about existing modal, spectral, transition-operator, sequence-analysis, qualitative, or mixed-methods practices, their prevalence, or the candidate's novelty relative to them."}