Policy on revisions and corrections to published statistics
First published: April 2019
Converted to a webpage: October 2026
Background
This policy sets out SEPA’s policy for responding to scheduled revisions and unscheduled corrections of published statistics (including, but not limited to, Official Statistics). It has been developed in line with principles set out in the Code of Practice for Statistics (‘the Code’) published by the Office for Statistics Regulation. The Code specifies that revisions and corrections of errors should be released transparently and as soon as possible in line with a published policy, being clear about the nature and scale of change. Our aim is to manage revisions and corrections in a way that is timely yet minimises inconvenience to the data user, and protects confidence and trust in SEPA and in the statistics.
This policy covers changes to statistics that are prepared and published by SEPA, typically on a quarterly or annual basis. It does not cover published raw or ‘live’ data which can be updated at any time by the data provider. Although this policy is designed to support Official Statistics, it is applied to all statistics published by SEPA.
Regular changes to statistics can be inconvenient to users and can undermine confidence and trust in both the statistics themselves and in the organisation that publishes them. They may also adversely affect policy decisions and debate. For this reason our position for scheduled revisions and our preferred position for unscheduled corrections is to update published statistics once a year, usually at the time of the next publication. For official statistics, this policy applies from the pre-release stage onwards. In other words, once the official statistic is pre-released for publication, there will usually be no further changes through the year until the next publication date.
Scheduled revisions
Scheduled revisions usually arise from improvements in statistical methods, the availability of more and fuller data, or changes in policy that impact definitions. Our approach to scheduled revisions is as follows:
We will update the statistic at the next publication date.
We will identify, as far as possible, the key users of the published statistic. Key users are generally those who we directly report the data to and who are known by us to use the data.
We will give as much notice as possible to known key users of scheduled revisions to methodology, definitions or any other changes likely to lead to revisions, along with an indication of their possible scale and nature. Commonly, this will be indicated in the quality report that accompanies the published statistic.
Where relevant, we will consult known key users before making changes that affect the published statistic (for example, to definitions or methods).
Scheduled revisions will be subject to the same arrangements for publication and pre-announcement as other releases.
We will provide a statement to known key users explaining the nature and extent of revisions at the same time as they are released. This statement will usually be included with the next statistical publication.
Ideally, previous versions of the published statistic containing unrevised figures will be replaced and archived offline. This may not always be feasible. As a minimum all unrevised figures will be clearly marked and contain a link to the revised figures.
We will ensure that where time series are revised, or changes are made to methods or coverage, we produce consistent historical data, wherever possible, and user guidance. If it is not possible to produce consistent historical data then advice on potential effects will be given where possible.
Unscheduled corrections
SEPA’s systems and procedures are designed to minimise the risk of unscheduled corrections. Nevertheless, sometimes there are unexpected changes to the source data or errors in the methods used to produce the statistic. We use the following approach to arrive at a proportionate response to unscheduled corrections:
We will ensure that we have ascertained the full nature, extent and scale of the correction, consulting technical experts as required to help with the assessment.
An appropriate assessment and response will be based on a consideration of the following matters:
The importance of the correction, e.g. does it affect a headline figure?
The magnitude of the correction, particularly relative to any stated levels of accuracy.
The recipients of the correction - who is affected and how?
The consequences of the correction - does it alter the main message of the published statistic, or any likely use of it?
We will record the results of the assessment and the process used to make it.
Minor corrections will be corrected in the next edition of the publication. The correction will be made clear and the reasons explained. This may be contained in the next statistical publication.
For major corrections we will:
Define an appropriate timescale for corrections to be completed.
Update the published statistic, make clear this has been done, and give an indication of the nature and scale of the correction.
Notify known key users, especially where we believe the correction could affect their own work.
If relevant, issue a new or updated formal Statistical News Release subject to the normal procedures for such releases.
Lessons learned
In an effort to learn from experience and minimise the need for revisions and corrections, we will do the following:
Review and, if appropriate, amend our procedures and methods.
Work with data providers as appropriate.
Consider publishing an explanation of the above actions where this would increase confidence in the data and avoid the need for future change.
Contact
If you would like more information about this policy, you can:
telephone us on 0300 099 66 99
Glossary of terms
Official statistics; Statistics produced by crown bodies, those acting on behalf of crown bodies, or those specified in statutory orders, as defined in Section 6 of the Statistics and Registration Service Act 2007.
Scheduled revisions; Planned amendments to published statistics in order to improve quality by incorporating additional data that were unavailable at the point of initial publication.
Statistics; A collection of measures about a particular attribute compiled from a set of data. Statistics are used for making generalisations or inferring conclusions about particular attributes, at an aggregate level, for example, about a particular subset of the population.
Unscheduled corrections; Amendments made to published statistics in response to the identification of errors or changes following their initial publication.