Skip to content
PlagiatScanner.de
Examination procedures · action · PRV-05

Use version history as evidence of your own work

In brief: Preserve native history and original files, select meaningful milestones and explain the particular changes between them. Connect those states with research notes, sources, feedback or tests. Versions can make the development of a paper plausible, but they do not establish the identity of the person writing or the completeness of the workflow on their own. Any history reconstructed later must retain its real date and an explicit label.

Version types and evidential valueReviewed 11 September 2026

Different version systems preserve different parts of a workflow

A cloud document may hold numerous automatic states, while local files record only deliberate saves. Git commits are useful for prose, code and small reviewable changes, but only if the repository was genuinely maintained during the project. A PDF freezes a readable rendering yet normally loses editing events. An email attachment can support that a file existed by transmission time, not how every part was produced.

Inventory the real locations first: word processor, cloud storage, local disk, version-control service, email, learning platform and backups. Use native export functions without editing existing states. Record the provider, project or account context, export date and fields retained. Keep confidential records only in authorised storage.

Authority asset: version-type evidence map

What common version traces can and cannot show
Version typeUseful contributionImportant limitationHelpful companion
Native document historySequence of visible edits within the serviceAn account does not prove who made every editExport context and intellectual milestones
Local intermediate fileSpecific text state and file structureCopying can alter timestampsNote, backup or transmission record
Git commitLine change, sequence and commit contextAuthor fields are configurable; uncommitted work is absentReview, issue or test record
PDF or DOCX exportReadable rendering of a milestoneEditing events and external inputs are missingNative source and export note
Email attachmentFile existed in that state by the transmission timeInitial creation and editor remain openCorrespondence and earlier version
Reconstructed comparisonExplains differences observable todayNot a contemporaneous historyReal creation date and source list

Select intellectual milestones rather than flooding the reviewer

Choose states where a scholarly decision becomes visible: first viable outline, literature selection, method commitment, chapter draft, response to feedback, source audit and submission. An autosave every two minutes does not automatically improve understanding. Your selection must nevertheless be transparent; do not omit a relevant state because it conflicts with your account.

Assign neutral IDs such as V-01 to V-08. For each, record a date or period, storage location, chapter coverage, major open questions and the reason for the next change. A screenshot of cloud history is usually only supporting material. Where possible, preserve the native record or official export and describe what it includes.

Explain development through specific examples

A useful package shows more than file existence. Select a few representative changes. Connect an early research question with annotations, a method revision with supervisor feedback, and a rewritten result with an analysis record. Include abandoned directions when they are relevant to the concern.

In text comparison, distinguish insertion, deletion, movement and formatting. A large passage appearing in one step requires context: did it come from another file you wrote, a transcript, code output, permitted feedback or an AI response? Link the originating record. “I wrote it myself” does not explain an otherwise opaque jump.

For software and data work, executable tests, notebooks, issues and review comments may provide valuable companion evidence because they reveal decisions and corrections. The guide to documenting research file versions helps establish a reliable routine while a project is active.

Do not confuse technical metadata with proof of personal identity

A filesystem date may record a copy rather than the creation of content. A cloud username associates an event with an account, not necessarily the individual at the keyboard. Git author fields are configurable. A checksum verifies byte equality from a recorded reference event, not authorship or completeness before that event.

Use bounded language: “This version contains …”, “The service attributes the edit to account …” or “The attachment was transmitted on …”. Avoid “conclusively proves” where the record addresses only one aspect. Independent materials such as a version, contemporaneous feedback and aligned research notes can produce a stronger workflow account together.

Assemble a version bundle another person can navigate

The cover sheet names the work, submitted state, purpose and compilation date. Follow it with a one-page version index, selected comparisons and an appendix register. Every statement cites an ID. Retain native files alongside readable exports. Review third-party comments, personal data and confidential research before transmission.

  1. Identify the submitted file and earliest relevant state.
  2. Record the export route and retained metadata.
  3. Select milestones for stated reasons.
  4. Connect large changes with their genuine source records.
  5. Expose gaps, contradictions and later reconstructions.
  6. Limit technical claims to what each record supports.
  7. Preserve the bundle sent and its transmission receipt.

The examination procedures hub places version evidence in context. The AI-concern evidence inventory covers other record types. The one foundation connection on this page is academic writing and evidence.

Official foundations

  1. German Research Foundation: Guidelines for Safeguarding Good Research Practice – traceable documentation of research processes.
  2. ALLEA: The European Code of Conduct for Research Integrity – reliable records, transparency and reproducibility.
  3. German Research Foundation: Research Data – official guidance context for responsible data management and reuse.

Sources reviewed 4 September 2026. Institutional requirements may specify additional formats or retention duties.