{"actors":["Sport-science principal investigator who defines the study claims","Biostatistician who specifies multiplicity controls","Research analysts who process physiological and performance data","Institutional methods reviewer who assigns evidence status","Coaches and practitioners who may later interpret the published recovery claim","Study participants whose data and training exposure require protection"],"affected_objective":"Ensure that a recovery-method claim from a multi-endpoint sport-science study receives an evidence status reflecting all measured outcomes, time points, stopping looks, and analytic alternatives before it is communicated as practice-ready.","arm":"ORDINARY_DIVERSE_P2","authority_safety":{"authorized_first_step":"The principal investigator and biostatistician may replay the proposed discipline on one completed, unpublished recovery study in shadow mode, using existing authorized data and producing only an internal comparison of claim statuses.","decision_authority":"The principal investigator retains responsibility for scientific interpretation, the institutional methods reviewer assigns status under the pilot rule, coaches retain ordinary practice authority, and participant-safety decisions remain with the study's designated clinical or safety lead.","excluded_actions":["Changing participant training, treatment, recovery procedures, eligibility, or compensation during the shadow replay","Submitting, withdrawing, correcting, or retracting a manuscript on the basis of the pilot without a separate authorized review","Presenting exploratory or multiplicity-adjusted findings as independently replicated","Changing endpoint families or decision thresholds after viewing the replay results","Recruiting participants or collecting new data during the first step","Expanding access to identifiable participant data"],"halt_rollback":"Halt the replay if the original analysis history cannot be separated from reconstructed assumptions, access controls fail, or the review delays an active participant-safety response. Delete temporary derived pilot outputs under the study's data-retention rules, retain an access-controlled methods note, and return manuscript decisions to the pre-pilot process."},"baseline":"In the inferred baseline, a recovery-method study measures several physiological, perceptual, and performance outcomes at multiple post-exercise time points, permits alternative exclusions and model specifications, and may inspect results before completing enrollment. The manuscript foregrounds the most favorable comparison using its ordinary single-test interpretation, while the full opportunity set and abandoned analyses are not carried into the claim's evidence status.","candidate_id":"multiple_testing_discipline__sport_science__ORDINARY_DIVERSE_P2","causal_chain":["A study comparing recovery methods creates many claim opportunities across biomarkers, soreness ratings, strength tests, sprint measures, time points, contrasts, exclusions, models, and interim looks.","Chance variation can make at least one method-outcome-time combination appear unusually favorable even when individual comparisons use an ordinary threshold.","Selecting that combination for the title or conclusion while omitting the unsuccessful opportunities makes it appear closer to a planned confirmatory result.","The selected claim can then shape subsequent research priorities or practitioner interpretation despite arising from an incompletely represented search process.","The intervention defines endpoint and analysis families before confirmatory evaluation and inventories every permitted and attempted look.","A hierarchical familywise rule allocates the error budget across the primary recovery claim and subordinate endpoint families, while unplanned analyses remain explicitly exploratory.","A multiverse report exposes whether the selected conclusion depends on favorable exclusions, transformations, or model choices rather than silently choosing one path.","Only a separately preregistered replication using newly collected participant data can advance the selected claim from adjusted discovery to replicated and practice-review eligible.","A retained evidence passport connects every reported statement to its family, analytic history, status, and replication obligation."],"cell_id":"multiple_testing_discipline__sport_science","consequence":"A chance-selected recovery result could be communicated as a general method advantage, directing practitioner attention and follow-up research toward a claim whose apparent evidentiary strength does not reflect the number of opportunities that generated it.","diversity_from_prior_proposals":"The sealed P1 concerns weekly athlete-monitoring searches that can trigger individual workload restrictions. This opportunity instead concerns multi-endpoint recovery experiments, manuscript-level claims, hierarchical familywise testing, analytic-path disclosure, and independent replication before practice-review eligibility; it neither modifies nor repairs the P1 workflow.","experiment_id":"eoa_inverse_innovation_exp13_second_slot_policy60_20260806","intervention":"Create a Recovery Study Evidence Passport for multi-endpoint experiments. Before confirmatory analysis, the investigator registers the overarching recovery claim, its primary outcome and time point, subordinate physiological, perceptual, and performance families, allowed contrasts, enrollment looks, exclusions, transformations, and models. A controlled analysis workspace appends every executed or inspected alternative to the passport. A predeclared hierarchical familywise procedure governs confirmatory claims, with an alpha-spending schedule if enrollment is viewed sequentially; analyses outside the hierarchy are retained as exploratory. The manuscript must pair each conclusion with a status of exploratory, adjusted-discovery, confirmed, replicated, or rejected and include a multiverse summary for materially plausible analytic paths. A selected non-primary finding cannot become practice-review eligible until a separate team or firewalled analyst preregisters and tests the fixed claim in newly collected participant data.","mechanism_mapping":[{"counterfactual_removal":"Without advance registration, investigators could redefine endpoint families, models, or confirmatory priorities after observing favorable recovery results.","mechanism_slug":"preregistration","role":"Fixes the primary claim, subordinate families, eligible analyses, sequential-look schedule, and status rules before confirmatory results are examined."},{"counterfactual_removal":"Without a metric hierarchy, a favorable secondary biomarker or time point could replace a disappointing primary recovery outcome without an evidentiary penalty.","mechanism_slug":"metric_hierarchy","role":"Orders the primary recovery claim and subordinate physiological, perceptual, and performance families for confirmatory interpretation."},{"counterfactual_removal":"Without familywise control, the ordinary threshold for each endpoint would ignore the collective opportunity for at least one favorable recovery claim.","mechanism_slug":"bonferroni_like_correction","role":"Implements the predeclared hierarchical familywise error allocation across confirmatory endpoint families."},{"counterfactual_removal":"Without a sequential-look budget, repeated enrollment checks would create uncounted chances to stop when a favorable result appears.","mechanism_slug":"alpha_spending_plan","role":"Accounts for any permitted interim analyses of accumulating participant data."},{"counterfactual_removal":"Without the multiverse summary, dependence on one favorable exclusion, transformation, or model could remain hidden behind a single reported estimate.","mechanism_slug":"multiverse_analysis_report","role":"Displays the selected finding across registered, materially plausible analytic paths and preserves null or contrary alternatives."},{"counterfactual_removal":"Without an independent new-data test, a finding selected from the original endpoint and analysis search could be promoted using evidence shaped by that same search.","mechanism_slug":"replication_study","role":"Requires a fixed-claim test on newly collected participant data before practice-review eligibility."}],"nearest_rivals":["A single-hypothesis testing frame would fit the registered primary endpoint alone but would not account for the broader endpoint, time-point, model, and interim-look opportunity set.","A reproducibility protocol could rerun the reported recovery analysis without changing the evidentiary status created by selecting it from many alternatives.","Effect-size reporting could address the magnitude and practical interpretation of a recovery contrast after estimation but would not account for its selection from many claims.","Power-aware design could address sensitivity to a prespecified recovery effect but would not control false-positive opportunities across the study's many outcomes and analytic paths.","Confounder control could address distorted recovery comparisons but would not restore the missing multiplicity context of a selected result."],"negative_tests":{"intervention_falsifier":"The passport is operationally falsified if a reviewer cannot reconstruct all executed endpoint, time-point, exclusion, model, and interim looks; if analysts can generate manuscript-eligible findings through unlogged paths; or if claim families and error allocations change after results are known.","problem_falsifier":"The inferred problem is absent if records show one prospectively fixed recovery claim, outcome, time point, contrast, model, exclusion rule, and final analysis, with no interim inspection or additional analysis capable of supporting the same conclusion.","risks":["Investigators may define endpoint families too narrowly to reduce the apparent multiplicity burden.","Analyses conducted in personal software or informal spreadsheets may escape the passport.","A strict familywise hierarchy may classify potentially useful secondary findings only as exploratory.","Analytic-path reporting may become lengthy without improving how conclusions are labeled.","Replication analysts may be influenced by detailed knowledge of the original search.","Administrative requirements may delay manuscript preparation.","Multiplicity discipline may be mistaken for evidence of practical importance, sound measurement, causal identification, or participant safety."],"strongest_counterevidence":"The strongest counterevidence would be timestamped records demonstrating that recovery-study conclusions already arise from prospectively fixed endpoint families and analysis paths, account for every interim look, preserve all attempted alternatives, and require independent new-data replication before practice-facing communication."},"next_evidence_step":"Apply the passport retrospectively in shadow mode to one completed, unpublished recovery study. Reconstruct the endpoint families and attempted analysis inventory from protocols, code, outputs, and timestamps; separately record any element that cannot be verified. Then compare the manuscript's existing claim labels with those produced by the hierarchical familywise, sequential-look, multiverse, and replication-status rules. Have an independent methods reviewer determine whether the opportunity set and status transitions are reconstructable. Do not alter the study, manuscript, participant records, or practitioner guidance during this step.","observable_state":"The evidence passport displays the registered recovery claim hierarchy; every physiological, perceptual, and performance endpoint; sampled time points; treatment contrasts; exclusions; transformations; model specifications; interim looks; executed analyses; null and selected results; allocated error budget; multiverse summary; current claim status; reviewer; and any unmet replication requirement. Timestamps distinguish prospective specifications from post-result additions.","prior_art_status":"UNSEARCHED","problem":"A sport-science recovery experiment may compare several modalities across many biomarkers, subjective measures, performance tests, post-exercise time points, exclusion rules, models, and enrollment looks, then foreground the most favorable result as though it were the study's sole planned claim. The missing search context can give a selected recovery finding an unwarranted confirmatory or practice-facing status.","proposal_index":2,"remaining_contrastive_claim":"A recovery-method conclusion selected from multiple endpoints, time points, interim looks, and analytic paths cannot become practice-review eligible through an isolated single-test interpretation; it must carry the full study opportunity set, satisfy the registered family-level rule, and survive an independent fixed-claim replication.","revision_record":{"claim_changes":[],"conceptual_changes":[],"evidence_changes":[],"operational_changes":[],"parent_version":null,"progress_targets_addressed":["Defined a manuscript-level sport-science problem distinct from athlete load management","Preserved the claim-family, attempted-look, error-policy, status, confirmation, and discovery-record structure","Specified hierarchical familywise and sequential-look rules suited to the proposed study context","Separated exploratory analysis from independent replication","Bounded the first step to a reversible shadow replay with no participant-facing or publication action","Included explicit problem and intervention falsifiers"]},"schema_version":1,"structural_mapping":[{"archetype_element":"Define the claim family","domain_realization":"The passport groups the primary recovery claim and subordinate physiological, perceptual, and performance claims across modalities, time points, contrasts, and eligible analysis paths."},{"archetype_element":"Inventory attempted looks","domain_realization":"The controlled workspace records every executed endpoint analysis, interim inspection, exclusion, transformation, model, and abandoned alternative."},{"archetype_element":"Set the error-risk policy","domain_realization":"Confirmatory manuscript claims use familywise protection, exploratory leads remain available for follow-up, and practice-review eligibility requires independent replication."},{"archetype_element":"Apply a multiplicity-aware rule","domain_realization":"A hierarchical familywise procedure allocates the error budget across endpoint families, while an alpha-spending schedule accounts for registered interim looks."},{"archetype_element":"Label claim status","domain_realization":"Each manuscript claim is labeled exploratory, adjusted-discovery, confirmed, replicated, or rejected and linked to its unmet requirements."},{"archetype_element":"Confirm before costly action","domain_realization":"A selected non-primary recovery finding requires a separately preregistered test in newly collected participant data before practice-review eligibility."},{"archetype_element":"Preserve the discovery record","domain_realization":"The evidence passport retains specifications, timestamps, all attempted analyses, null results, analytic alternatives, status changes, reviewers, and replication obligations."}],"title":"Recovery Study Evidence Passport","version":0}