Guidance

Statement on revision policy

Updated 14 January 2022

Scope and purpose

This document sets out how revisions to the Department for Transport’s statistics outputs, whether scheduled or not, take place in line with the Code of Practice for Statistics.

Statistics are by their nature subject to error and uncertainty. Initial estimates are often systematically amended to reflect more complete information. Improvements in methodologies and systems can help to make revised series more accurate and more useful.

The department aims to avoid the need for revisions to statistics but has put in place the following policy to make any necessary revisions transparent to users.

Policy

A revision is a change to any official statistic once it has been placed in the public domain, whether through a printed document or electronically.

Scheduled revisions

Scheduled revisions are planned amendments to published statistics and a routine part of the production of official statistics. They will be managed systematically, reflected in dissemination plans and pre-announced in the release calendar in line with the Code of Practice. Such revisions will take place at the next scheduled publication of the series.

Scheduled revisions can occur for several reasons. They can occur when they form a normal part of the release process for a given series. This is common in highly aggregate series which are often first released in a less than complete state in order to meet the needs of users for timely information. Results based on more complete responses and analyses are then released at a later date. These statistics are initially marked as provisional to reflect the later scheduled revisions.

Scheduled revisions can also occur as a result of seasonal adjustment, benchmarking or rebasing exercises. Examples include regular updates to take account of further supplies of data, rebasing of an index to take account of new information or annual surveys replacing the results of quarterly or monthly surveys.

Scheduled revisions may also be due to changes, for example, to the underlying methodology, or to sources or classifications, or changes to geographical boundaries (such as transport authority or local authority).

Unscheduled revisions

Unscheduled revisions are those which do not fit into the managed pattern of revisions normally associated with the statistics in question. They can occur, for instance, at an unusual point in the publication cycle, or have an impact on the statistics which was not anticipated.

They can be caused, for example, by errors, by unforeseen changes in methodology, by the unexpected acquisition of new data, or by the effects of changes that may be made, for both statistical and non-statistical reasons, to any administrative or management systems that contribute data to official statistics.

Unscheduled revisions can either be minor revisions or substantial revisions.

Minor revisions

Minor revisions are revisions that do not change the context or misreport headline figures within reasonable rounding.

For those revisions deemed by the Head of Profession to be minor, a replacement document will be issued clearly marked by the date of the revision with a note giving a reason for the revision, identifying the statistics that have been affected and explaining clearly the scale, nature, cause and impact of the changes in the statistics Code of Practice, Q3.4. Where results are affected by methodological changes, a description of the change will be given.

If the error is minor and does not change the context or misreport headline figures within reasonable rounding, the figures will be revised with the next publication.

When unscheduled revisions are needed, they will take place as soon as is practicable Code of Practice, T3.9. This will be a decision from the Head of Profession (HoP), as to whether it will be corrected as soon as possible or as part of the next publication.

Substantial revisions

Substantial revisions, whether ‘scheduled’ or not, are those which lie outside the range of revisions normally associated with the statistics in question and which tend, therefore, to have a more significant impact.

Decisions on the content, format, and timing of releases which incorporate substantial revisions will be made by the Head of Profession. It is substantial if it changes the story or overall trend.

At the minimum, the statistics publication will be either revised as soon as possible and accompanied by a note to explain the scale, nature, cause and impact of the changes. Or, if work to correct the statistics would take longer, the statistics publication will be removed from the department’s website and replaced by a statement giving a reason for the revision and a timetable for the replacement statistics at the same time as placing a notice on the department’s statistics release calendar.