- Published
- Updated
When result tables change during drafting, the abstract, text and figures may rely on different versions. File-modification dates may not reliably explain which result set entered scientific review or which numbers authors intended to approve.
The publication decision
Define the publication table set through a version and data cut. Freezing does not prohibit necessary corrections; it provides a shared baseline for managing their effects visibly. Mark tables as final, provisional or awaiting clarification. Support writers in using controlled sources rather than copying isolated numbers from email exchanges that lack a clear relationship to the accepted analysis.
Worked example
For example, a safety table is revised while the abstract retains an earlier event count. The team links changed cells to an impact list and checks text, figures, footnotes and abstract together. It distributes the revised set as a new version and assesses which content earlier author comments and approvals actually covered.
Coordination checks
- Link the table-set version to its data cut and analysis scope.
- Trace changed results to every use within the manuscript.
- Prevent provisional tables being presented as final findings.
Acceptance and boundaries
Acceptance means the manuscript’s numbers rely on one explainable result set. This consistency control does not itself prove statistical correctness, which needs relevant expert review. If a frozen table contains an error, make a reasoned update and assess affected approvals. Hidden corrections or replacing files without version changes weaken the ability to establish consistency later.
Frequently asked questions
- Does freezing mean changes are prohibited?
- No. Necessary changes can proceed through controlled versioning and impact assessment.
- Is manually correcting the number in the text enough?
- Check its other uses in the abstract, figures and footnotes as well.