Transparency data

Corrections and revisions policy for official statistics

Updated 8 March 2023

Applies to England

1. Policy statement

We aim to minimise revisions to our statistical publications but sometimes changes are unavoidable. These changes come in two forms:

1. Corrections – these are unplanned and occur when errors in either the data or the methodology are found after release of the publication.
2. Revisions – these are generally, but not always, planned, and are the result either of improvements in methods or of additional data.

We will list any corrections or revisions on our website.

This policy follows the best practice set out in the Code of Practice for Official Statistics

2. Corrections

We have robust procedures in place to minimise errors and regularly review them. However, errors do sometimes occur.

When errors occur we are open and transparent about the need for any corrections. We assess the extent to which the error could affect interpretation of the publication and take appropriate action. If an error which could affect interpretation is found after a release has been published the report will be republished.

Once published, data are not usually subject to revision, although subsequent releases may be revised to insert late data or to correct an error.

3. Revisions

3.1 Methodological changes

If changes of the following type are to occur, we will, where possible, notify users in advance on our website and consult with them via user groups:

• major changes to collection systems
• changes in statistical methodology, definitions or categories

We strive to maintain a consistent time-series and will consider all of the available options that would enable us to do that (including any revisions to previously published time-series).

3.2 Communication of revisions

The statistical release page for each product will have a link to the revisions policy for that publication. The link will contain:

• a note that identifies any revisions made to that product and the significance of those changes
• any indication of the status of the figures – this would indicate whether they are usually subject to revision and in what way
• a link to a page explaining the impact of any revision if it would result in reinterpretation of economic or social developments

4. General issues

4.1 Pre-release conditions applied to changes

Some individuals receive statistical publications before they are published on our website. Where a change is needed it may be subject to the same pre-release conditions as new statistics. We will consult Ofqual’s Head of Profession for Statistics before we publish any change that would have a major economic or political impact.

4.2 Timeliness of changes

Timeliness is a measure of the gap between the date when the data was collected and the date when the final report was published. We try to minimise this gap and publish all reports at the earliest opportunity. We will balance the need for timeliness against the need to avoid frequent changes.

4.3 Keeping a history of changes

We keep electronic records of all previous versions of statistics where possible, which are available to users upon request where practicable. To avoid confusion, we only keep the most up-to-date report on our website.

4.4 Analysis of changes

We monitor the changes that we make in order to improve our statistical production process and the quality of our statistics, and to identify any systematic bias. If we find any such bias, we will notify our users.