Returned-result correction workspace

Correction history for returned laboratory results

A browser-based research-use workspace for preserving the received predecessor, recording a proposed correction, exposing affected downstream objects and routing scientific impact, notification and rerelease decisions to accountable people.

SingularCell Research. Reviewed for primary-source scope, version-history semantics and unsupported compliance or scientific claims. The workspace records supplied history and review states; it does not determine truth, impact, adequacy, acceptance or regulatory applicability.

How to use this resource

Use synthetic or non-confidential metadata only. Record one primary event plus every applicable change dimension, bind exact versions, keep unknown evidence visible and export the draft into your controlled review system. This page does not determine whether a correction is true, scientifically material, adequate or acceptable.

Entries stay in this browser page and are not submitted to SingularCell. Download or storage is not implied.

Twelve event types that must not collapse

Choose a primary event, then attach every applicable change dimension or linked component. The labels describe record history; they do not decide truth, cause, impact or adequacy.

Choose a primary event, then attach every applicable change dimension or linked component. The labels describe record history; they do not decide truth, cause, impact or adequacy.
Event type	Use when	Do not infer
SOURCE_RECORD_CORRECTION	The supplying organization reports that one of its source records should change	The supplier is correct or the predecessor was false
METADATA_CORRECTION	Context, identity mapping, unit, label or descriptive metadata changes	Unchanged data bytes mean no scientific impact
DATA_VALUE_CORRECTION	One or more recorded observation values change	The successor value is scientifically correct
REPROCESSING_OR_REANALYSIS	Existing inputs are processed again with identified code, method, configuration or parameters	Source data changed or the new analysis is justified
REPLACEMENT_MEASUREMENT_OR_RERUN	New experimental or measurement activity produces new observations	The rerun repairs, erases or replaces the original measurement
REPORT_CORRECTION	A released report location, text, table or figure changes	Underlying observations changed
INTERPRETATION_UPDATE	A human conclusion or explanation changes while observations remain preserved	The updated interpretation rewrites the observations
WITHDRAWAL_OR_INVALIDATION	An object ceases to be current or available for a declared scope	Deletion, misconduct or universal invalidity
STUDY_PLAN_AMENDMENT	An intended plan change creates a successor study-plan version	The change was part of the original plan
DEVIATION	Execution departed unintentionally from the governing plan	The deviation automatically invalidates the study
ADMINISTRATIVE_CLARIFICATION	Non-substantive wording or administrative context is clarified	The label proves no scientific impact
REFORMAT_OR_TRANSPORT_TRANSFORMATION	Encoding, packaging or transport representation changes	Semantic content and scientific meaning are unchanged unless checked

Draft one bounded correction event

Transfer only approved, non-confidential information into the controlled system that owns the real study history. Blank fields remain unresolved, not assumed.

Case and version scope

Name the exact study, contract, returned package, affected object and current ruleset versions.

Primary event and change dimensions

Choose one primary event, then list every overlapping metadata, data, processing, report, interpretation or lifecycle dimension.

Predecessor and candidate successor

Record stable IDs, versions, schema or manifest references, digest status and availability without superseding the predecessor early.

Change and comparison scope

List old and new fields, files or report locations and the exact comparison method, inputs, time and limitations.

Reason, actor and supporting evidence

Separate the source's stated reason from independently established cause and record actor, role, organization, site, time and evidence references.

Downstream dependency and impact state

List known affected datasets, analyses, reports, reviews, exports and decisions plus graph gaps and required human review.

Review and rerelease state

Record separate technical, data, scientific, release and governance decisions bound to the exact candidate successor.

Notification and acknowledgement

Separate send, delivery, acknowledgement request, due time, receipt response and substantive acceptance for each recipient.

Downstream impact and rerelease checklist

A checked item means the named record exists for the exact correction case. It does not mean the scientific consequence is known or resolved.

Freeze the received predecessorPreserve the exact received package and affected object before recording the correction notice.Evidence prompt: Reference immutable predecessor IDs, versions, manifests and availability state.
Record a candidate successorLink the candidate to its predecessor without presenting it as released, current or scientifically correct.Evidence prompt: Reference revision relation, candidate state and supplied source assertion.
Record the scoped differenceState exactly what was compared, how, when and where comparison was impossible or restricted.Evidence prompt: Reference comparison inputs, method version, results and limitations.
Traverse known dependenciesFind datasets, analyses, figures, reports, passports, reviews, exports and decisions that reference the predecessor.Evidence prompt: Reference graph scope and retain DOWNSTREAM_DEPENDENCY_UNKNOWN for gaps.
Mark potentially stale objectsPrevent objects based on the predecessor from silently remaining current while impact is unresolved.Evidence prompt: Reference each object's version, relation and required next action.
Separate technical and scientific reviewKeep structural checks, data stewardship, scientific impact, release and governance decisions as separate scoped records.Evidence prompt: Reference named roles, exact subject version, scope and decision state.
Regenerate and rereview affected outputsCreate successors for affected derived objects and bind fresh review instead of silently carrying old approval forward.Evidence prompt: Reference new outputs, inputs, methods, reviewers and predecessor relationships.
Notify and track acknowledgementRecord each intended recipient, channel, send time, delivery status, due time, acknowledgement and acceptance separately.Evidence prompt: Reference delivery evidence and overdue or failed states without inferring acceptance.
Release, withhold, withdraw or close with limitsBind the final workflow action to exact versions, accountable decisions, unresolved items and stated limitations.Evidence prompt: Reference rerelease or withholding record and preserved predecessor history.

Lifecycle states without false assurance

Use precise states so progress through a workflow cannot be mistaken for scientific resolution.

Use precise states so progress through a workflow cannot be mistaken for scientific resolution.
State	Exact meaning	Does not establish
REPORTED_UNVERIFIED	A correction assertion was received and preserved	That the assertion is true or complete
PREDECESSOR_PRESERVED	The referenced prior version remains retrievable in the stated scope	Original authorship, custody or accuracy
SUCCESSOR_RECEIVED	A candidate successor and relation were recorded	That the successor is current, correct or acceptable
DIFF_RECORDED	Named locations or bytes were compared under an identified method	Semantic completeness or scientific impact
IMPACT_REVIEW_REQUIRED	Known dependencies or uncertainty require accountable review	That impact exists, is material or is resolved
SCIENTIFIC_REVIEW_COMPLETE	A named reviewer recorded a bounded decision for an exact version	Universal validity, compliance or truth
APPROVED_FOR_RERELEASE	A scoped release decision exists for the exact successor	Regulatory acceptance or scientific proof
RERELEASED	The successor was made current for the recorded scope	That every recipient received or accepted it
CLOSED_WITH_UNRESOLVED_ITEMS	Workflow activity ended while named issues remain visible	That the correction is scientifically resolved

Synthetic correction-event object

This deliberately incomplete object preserves a laboratory assertion, keeps the predecessor current with an alert, identifies a candidate successor and routes known dependencies for review.

Synthetic research-use correction event

{
  "example_classification": "SYNTHETIC",
  "correction_event_id": "corr:study-1042:01",
  "study_contract_version_id": "study-contract:1042:v3",
  "received_package_version_id": "result-package:1042:received-v1",
  "primary_event_type": "SOURCE_RECORD_CORRECTION",
  "change_dimensions": ["METADATA_CORRECTION"],
  "status": "IMPACT_REVIEW_REQUIRED",
  "source_assertion_status": "SUPPLIER_DECLARED_NOT_INDEPENDENTLY_VERIFIED",
  "predecessor": {
    "version_id": "sample-manifest:1042:v1",
    "hash_status": "SYNTHETIC_PLACEHOLDER_NOT_COMPUTED",
    "availability_status": "CURRENT_WITH_CORRECTION_ALERT"
  },
  "candidate_successor": {
    "version_id": "sample-manifest:1042:v2-candidate",
    "was_revision_of": "sample-manifest:1042:v1",
    "hash_status": "SYNTHETIC_PLACEHOLDER_NOT_COMPUTED",
    "release_status": "CANDIDATE_NOT_RELEASED"
  },
  "changes": [{
    "location": "/rows/17/well_id",
    "old_value": "B04",
    "new_value": "B05",
    "comparison_method": "FIELD_DIFF_V1"
  }],
  "downstream_impacts": [{
    "object_id": "report:1042:v1",
    "status": "POTENTIALLY_STALE",
    "required_action": "SCIENTIFIC_IMPACT_REVIEW"
  }],
  "notification": {
    "recipient_role": "SPONSOR_STUDY_OWNER",
    "sent_at": "2026-08-29T02:20:00Z",
    "delivery_status": "SENT_DELIVERY_UNCONFIRMED",
    "acknowledgement_due_at": "2026-08-30T02:20:00Z",
    "acknowledgement_status": "ACKNOWLEDGEMENT_NOT_YET_ASSESSABLE"
  },
  "limitations": [
    "Supplier assertion only; biological correctness is not established.",
    "Dependency detection does not determine scientific materiality."
  ]
}

Boundary: This synthetic object demonstrates record structure only. It does not establish that the correction is true, complete, scientifically adequate, accepted, released or compliant.

Never silently replace a returned laboratory result. Preserve the received version, create a linked correction event and candidate successor, identify potentially stale downstream objects, and require scoped human review before rerelease.

Preserve what was received before describing what changed

A correction history begins with an exact predecessor: the result package, dataset, manifest, analysis, report or interpretation that was received. Do not edit that object in place. Record its stable ID, version, available manifest or digest status, schema, source assertion and current availability before attaching a notice or candidate successor.

A proposed successor does not make the predecessor superseded. Until an accountable release decision occurs, the predecessor should remain current with a visible correction alert and the successor should remain candidate-not-released. Ordinary supersession is a revision relationship and application state; W3C invalidation should be reserved for actual cessation or expiry within the asserted scope.

Exact study, contract, returned-package and affected-object versions
Received predecessor preserved without silent overwrite
Candidate successor identified separately from the current release
Source assertion distinguished from independently established fact
Hash or manifest state reported as checked, missing, mismatched or not computed

Describe the event without deciding its scientific meaning

Correction labels can overlap. A laboratory notice may be a source-record correction whose changed dimension is metadata or data; reprocessing may generate a derived successor without altering source observations. Record one primary event and one or more change dimensions, or link multiple typed events, rather than forcing an issue into one mutually exclusive label.

Deterministic software can compare declared fields, files, manifests and versions and traverse recorded dependencies. It cannot decide whether the source explanation is true, whether a changed mapping is scientifically material, whether a rerun was justified or whether a corrected result should be accepted. Those decisions need named people, exact scope and evidence.

Primary event and every applicable change dimension
Old and new fields, files or report locations with comparison scope
Reason and root-cause status kept separate
Actor, role, organization, site and event time
Technical, data, scientific, release and governance review states

Expose downstream impact and communication state

A changed input can make datasets, analyses, figures, reports, Evidence Passports, reviews, exports and recorded decisions potentially stale. Dependency traversal identifies known paths; it does not prove that the graph is complete or decide scientific materiality. Unknown coverage must remain downstream-dependency-unknown rather than becoming no impact.

Communication history must separate sending, delivery, acknowledgement request, response deadline, receipt acknowledgement and substantive acceptance. A sent notice is not proof of delivery; acknowledgement is not acceptance. Rerelease should bind new technical, scientific and release decisions to the exact successor version while preserving every predecessor and unresolved item.

Known downstream objects and unresolved graph coverage
Potentially stale, regenerated, rereviewed and rereleased states
Notification channel, send time, delivery state and deadline
Receipt acknowledgement separated from substantive acceptance
Case closure allowed to retain unresolved issues and limitations

Primary sources

W3C — PROV Data Model ↗World Wide Web Consortium · Entities, activities, agents, revision, derivation and invalidation relationships for explicit predecessor, activity and successor history.
NIST — Research Data Framework, Version 2.0 ↗National Institute of Standards and Technology · Research-data lifecycle concepts including original-copy assertions, versions, derivatives, timestamps, updates and removed-data recognition.
eCFR — 21 CFR 11.10 ↗United States Electronic Code of Federal Regulations · Within covered closed systems, time-stamped audit trails and controls that do not obscure previously recorded information.
FDA — Part 11 scope and application guidance ↗United States Food and Drug Administration · A bounded interpretation of Part 11 scope tied to predicate-rule records and reliance on electronic records.
eCFR — 21 CFR 58.130 ↗United States Electronic Code of Federal Regulations · Within covered FDA GLP scope, data-entry changes preserve the original and record reason, date and responsible person.
eCFR — 21 CFR 58.185 ↗United States Electronic Code of Federal Regulations · Within covered FDA GLP scope, final-report corrections or additions identify the affected part, reason and responsible signed and dated person.
eCFR — 21 CFR 58.120 ↗United States Electronic Code of Federal Regulations · Within covered FDA GLP scope, protocol changes are documented with reason, signature and date and retained with the protocol.
OECD — Principles and guidance for GLP compliance monitoring ↗Organisation for Economic Co-operation and Development · Within OECD GLP context, intended study-plan amendments are distinguished from unintended deviations.
OECD — Advisory Document No. 22 on GLP data integrity ↗Organisation for Economic Co-operation and Development · Within OECD GLP context, original entries, reasons, actors, times, invalidated data, repeated processing and audit-trail review remain visible.
FDA — Data Integrity and Compliance With Drug CGMP guidance ↗United States Food and Drug Administration · Within drug-CGMP scope, contextual metadata and audit trails can include reprocessing and change justification.

Limitations

This is a research-use traceability aid. It records supplied correction assertions, versions, differences, dependencies, reviews and notifications; it does not verify that the history is true, authentic or complete.
Absence of a correction record means only that none was found in the stated systems, versions, time range and access scope; it does not prove that no error or upstream correction exists.
Hashes and deterministic comparisons apply only to exact identified inputs and do not establish correctness, source identity, authorship, custody, intent or scientific validity.
Dependency traversal reports only indexed relationships and cannot determine whether the graph is complete or whether a change is scientifically material.
FDA and OECD material is bounded to its Part 11, GLP or CGMP context and does not make ordinary research-use records regulated or SingularCell compliant.
Only accountable qualified people can determine factual cause, scientific impact, adequacy, acceptance, withdrawal, rerelease, legal obligations and regulatory applicability.
Do not enter confidential, personal, patient, proprietary, regulated or customer study information into this public browser page.