Producing official statistics

Revisions and corrections policy

Purpose

This policy sets out the Scottish Government's principles and procedures for handling revisions to published statistics and correcting errors, in accordance with the Code of Practice for Statistics.

Revisions

Revisions to statistics can be necessary and valuable when they incorporate new or improved data. However, we recognise that frequent or unexpected revisions can be inconvenient to users and may undermine confidence in statistics.

Any revisions to statistics can be inconvenient to users and can undermine confidence and trust. However, in some cases the user need for timely figures makes it desirable to release statistics on a provisional basis before they can be finalised making subsequent revisions necessary. This is often done on a regular basis and perhaps each time a series is published but can also arise from an urgent requirement to publish an official statistic.

Regular scheduled revisions may occur due to:

• availability of more complete or final data (e.g., replacing provisional with final figures)
• annual benchmark revisions to incorporate updated weights or population estimates
• methodological improvements or refinements
• incorporation of additional data sources
• changes to definitions or classifications to maintain international comparability

For each statistical series subject to regular scheduled revisions, we publish a revisions statement explaining:

• why the revisions are made
• when revisions typically occur
• the likely scale and nature of revisions (where this can be estimated)
• the period over which historical data may be revised
• clear labelling: Figures expected to be revised are clearly marked as "provisional" with an indication of when revised figures will be published

We provide as much advance notice as possible of:

• changes to methodology, definitions, coverage or other factors likely to lead to revisions
• the likely scale and nature of such revisions
• plans to improve consistency of historical time series following revisions

Revisions are subject to the same publication and pre-announcement arrangements as other releases. Forthcoming revisions are included on the Scottish Government publications schedule. All registered users are notified by email. When revisions are published, we provide a statement explaining the nature and extent of revisions.

When time series are revised or changes are made to methods or coverage:

• we produce consistent historical data wherever possible
• we provide clear user guidance explaining the changes
• where possible, we archive previous versions of publications
• as a minimum, unrevised figures are clearly labelled with links to revised figures

Corrections

Despite rigorous quality assurance, errors sometimes occur in statistical outputs. The Code of Practice requires us to correct errors and alert stakeholders promptly.

Our approach to corrections is proportionate to the significance of the error and its potential impact on users. When an error is identified, we assess its significance by considering:

• importance of the statistic: is it a headline figure or key indicator? How widely is it used?
• size of the error: what is the magnitude relative to the stated precision of the figures?
• impact on message: does the error alter the main message or conclusions?
• timeliness: how recently was the statistic published?
• potential consequences: could the error have affected decisions or led to misunderstanding?
• if data is actively promoted to users?

Based on these factors, we categorise our response based on significance:

Minor errors

These are errors with no material impact on interpretation or use of the data. You should:

• correct it as soon as possible
• where correction effort is disproportionate to impact of error,  then ensure that it is corrected in the next scheduled publication
• note the correction in the methodology or quality report
• inform the Office of the Chief Statistician about the error and the correction

Moderate errors

These are errors with some impact but not affecting main messages. You should:

• publish a correction notice on the publication page
• update affected tables and documents, clearly marked as corrected
• notify regular users via email
• inform the Office of the Chief Statistician about the error and the correction
• include a note in the next statistical bulletin

Major errors

These are errors affecting headline figures or main messages. You should:

• publish a prominent correction notice immediately
• issue a corrected publication with clear indication of changes
• alert all known users via email
• notify the Chief Statistician and OSR
• consider issuing a media statement if the statistics had significant public attention

In addition we maintain clear records of:

• the nature of errors identified and corrections made
• the assessment process and decisions taken
• steps taken to prevent recurrence
• communications with users and stakeholders

Governance

Statistical branch heads are responsible for implementing this policy for their outputs, in consultation with senior statisticians and the Chief Statistician where appropriate.

Decisions on the classification of errors and appropriate responses should be made by the statistical team in consultation with senior statisticians. The Chief Statistician should be consulted for major errors or complex cases.

User Feedback

Users who identify potential errors in published statistics should contact the relevant statistical team (contact details are provided in each publication) or the Office of the Chief Statistician.

Alternatively, you can contact OSR by email or via the OSR website.

Review

This policy is reviewed annually to ensure it remains appropriate and aligned with the Code of Practice for Statistics.