Introduction

Analysis, Statistics and Research Branch (ASRB) within the Department for Infrastructure (DfI) is responsible for producing and disseminating a range of Official Statistics and Accredited Official Statistics relating to the Department’s travel, transport and planning functions. Statistics and research published by ASRB are available on the DfI website.

Background

ASRB staff are statisticians within NISRA and work in compliance with the Code of Practice for Statistics. The Code sets out the standards that producers of official statistics are expected to meet to ensure their work serves the public good, as required under Section 10 of the Statistics and Registration Service Act 2007. The Code is based on the three core principles of:

Trustworthiness;
Quality; and
Value.

Under the principle of Trustworthiness, Standard 3.9 of the Code states:

Release revisions and corrections of errors transparently and as soon as possible in line with the organisation’s published policy, being clear about the nature and scale of change.

This document outlines ASRB’s policy in the event of scheduled or unplanned revisions and the correction of errors.

Revisions

This statement is issued in accordance with the requirements set out under the Code of Practice for Statistics and sets out our intention to be open and transparent about any revisions we make to Official or Accredited Official Statistics and to ensure that users of statistics produced by us have easy access to comprehensive information about those revisions.

We endeavour to produce the best figures we can, in light of the available information and resources. However, improvements and consequential revisions can result as we continually strive to improve the methods and sources.

Some series are more subject to revisions, for example, when early or provisional estimates are produced. In such cases we will make it clear which Official or Accredited Official Statistics are provisional and are subject to a scheduled pattern of revision. When an entire release is made up of provisional statistics, the whole document is marked as provisional; when final figures are subsequently released then these are clearly marked as final.

Our methods will be kept under review to minimise the scale of revisions. Where there has been a revision to a table, this will be noted in an appropriate footnote to that table. Where a publication has been revised, only the most recent version will be available on the website, and the title will indicate that it is a revised publication. The rationale for any planned or unplanned revision will be outlined in the report.

Correction of Errors

Many of the revisions we make are a normal, unremarkable and inevitable feature of statistical life and users are able to absorb and plan for those revisions accordingly. They reflect the receipt of fuller and more complete information. However, some revisions are ‘unexpected’ in the sense that they are the consequence of errors or weaknesses in procedures or systems, or are the result of errors in source material. Regardless of whether the responsibility is ours or others, we will follow the procedures described below.

Announcement of impending corrections: We will be open and transparent about the need for any unscheduled corrections. Once the need for a correction that warrants customer notification is ascertained, we will announce our intention to issue corrections, and the planned date of issue.
Dissemination of corrections: Decisions relating to the dissemination of unscheduled corrections will be made by the Principal Statistician within the Department responsible for the affected statistical series. In general terms:
Web versions of releases/publications/tables/articles/etc.: As soon as possible after we ascertain that a correction is necessary and warranted, we will amend the latest electronic versions of any release, publication, infographic, table, article, etc. which contains the affected statistics or text, and re-populate the website as soon as possible with those amended versions. If the correction is minor, or insignificant, in the sense of being inconsequential and hardly noticeable, we will insert the necessary changes without alerting anyone. Alternatively, we may accumulate minor corrections and make an update at the time of the next scheduled web publication to avoid making too many frequent but minor changes.
Paper versions of releases/publications/tables: Unless there are compelling reasons, we will not attempt to recall/re-issue any paper versions of any release, publication or table etc. that contains the affected statistics or text if it has already been distributed. Paper copy provision of reports is limited to where strictly necessary, so few copies are distributed. We will, of course, ensure that when further paper copies are issued, they will include any corrections that have been incorporated in the electronic versions. However, if the error is substantial or significant, we will, where practicable, notify any recipients of paper versions by post, telephone or e-mail, and point them to the revised version available on the web. If the release or publication has a long shelf-life, we will consider re-issuing a revised paper version in exceptional circumstances. Otherwise, requests will be redirected to the electronic, online versions.
Methodological and Reclassification Changes: Where a substantial methodological change is planned or data are to be reclassified to a new definition, we will announce our intentions before the release of the statistics based on the new methods and/or definitions. Substantial methodological changes will only take place following consultation with users. For any key official or Accredited Official Statistics, we will provide background information and an explanation about the changes. In the event of a methodological or reclassification change, DfI will:

i.Publish a statement explaining that a reclassification has taken place and impacted upon statistical data.

ii.Produce and publish a statement on the DfI website, or in the affected publication, explaining any subsequent changes to methods and coverage.

iii.Produce and publish any impact analyses of the change to statistical data and make these available on the DfI website or in the affected publication. Sometimes, it can be some time before the full impact of a particular change can be fully quantified as it can require the collection and analysis of future data. In such cases, an attempt will still be made to give some guidance as to the likely impact of the change and this will be revisited as new data allows for further more detailed analysis.

Feedback

We welcome feedback from users on our Revisions and Errors policy. Please send any comments to ASRB@nisra.gov.uk

Contact Details

For further information contact Regina McGeown or Martin Devlin
Principal Statisticians ASRB ASRB@nisra.gov.uk

ASRB Revisions and Errors Policy