Guidance

FCDO statistics: statement on revisions policy

Published 19 December 2023

This document sets out the FCDO’s approach to revisions of its statistical outputs, in line with the Code of Practice for Statistics (‘the Code’), to support its trustworthiness pillar by providing transparency on the FCDO’s revisions policy. This policy details both scheduled and unscheduled revisions which can include: corrections, changes in methodology, and/or changes made due to receipt of further data.

The FCDO aims to ensure that all published data is as robust as possible, with quality assurance processes to minimise the frequency and scale of any unscheduled revisions. Where revisions are deemed necessary by the Chief Statistician/Head of Profession, these will commit to the following principles of the Code.

• trustworthiness T 3.9, orderly release: scheduled revisions or unscheduled corrections to the statistics and data should be released as soon as practicable. The changes should be handled transparently in line with a published policy
• quality Q 2.5, sound methods: producers of statistics and data should provide users with advance notice about changes to methods, explaining why the changes are being made. A consistent time series should be produced, with back series provided where possible. Users should be made aware of the nature and extent of the change
• quality Q 3.4, assured quality: scheduled revisions, or unscheduled corrections that result from errors, should be explained alongside the statistics, being clear on the scale, nature, cause and impact

Scheduled revisions

Scheduled revisions to statistical outputs may happen due to changes to methods or to data. These are planned amendments to published statistics which aim to improve the quality of the statistics. Generally, any change to statistics will be incorporated into the next scheduled publication for the statistical output. However, the Chief Statistician/Head of Profession may deem it necessary to revise and republish figures earlier depending on the impact of changes.

Method changes

Planned changes in statistical methods or source data to improve the accuracy of statistical outputs may lead to revisions. Where major changes to methods are proposed, we will look to consult users and announce these in advance, where possible.

Outputs affected by methodological changes will have a clear description of the change and an explanation on the scale, nature, cause and impact of the change, where possible.

Data changes

Statistical outputs may be published as provisional and subsequently updated when more data becomes available. This allows an early indication of the emerging picture and trend.

Updates to data are common and a regular part of the statistics production process. These can occur for a number of reasons, including (but not limited to) data lags between programmes starting to actually delivering outputs; or obtaining confirmation of data from partners whose reporting calendar may not sync to our reporting schedule.

Revisions incorporating the new data would normally take place in the next scheduled publication of the statistical output.

Unscheduled revisions

Unscheduled revisions may be a result of errors in the statistical process or unexpected changes to methodology or source data.

Errors

Whilst there are robust systems in place to minimise the risk of error in our statistical processes, we recognise that errors may occasionally occur. The Chief Statistician/Head of Profession will assess the impact of any correction on the interpretation of its statistical output and decide on the appropriate course of action for correcting it and alert our users.

Major errors which impact the interpretation of the published statistics will be corrected and revised immediately. A clear explanation of the cause of the error along with its impact, extent and effect will also be provided with the revision.

Minor errors judged to not materially impact the interpretation of published statistics will be revised in the next scheduled release of the statistics. Changes due to these errors will be similarly made clear to users.

Unexpected changes to data or methodologies

Revisions may occur due to unexpected changes to methodology or data. For example, unexpected changes to any administrative or management systems that contribute data, or the unexpected acquisition of new data. The Chief Statistician/Head of Profession will assess the impact of any such changes and decide on the appropriate course of action for revisions in these instances.

Statistical outputs

The FCDO publishes accredited official Statistics on International Development, and statistics on the outcomes of its activities around the world.

Detail on publication specific revision processes can be found in each individual publication.