The Annals of Actuarial Science (AAS) editorial team and the Institute and Faculty of Actuaries (IFoA) believe that research articles should contain sufficient information to allow others to understand, verify and replicate new findings and that sharing data can help to improve the impact of research. Therefore, wherever possible, we encourage authors to make replication materials such as data and software code available to readers alongside their article and without undue barriers to access.

We recognise that not all authors of Original Research or Review Papers will be able to make replication materials available and wish to be clear that this is not a requirement of publication in AAS. However, please note that full replication materials need to be provided for Actuarial Software submissions.

How to make replication materials available

As part of the submission process, authors need to notify the Editor of their intentions in relation to replication materials, by completing a Data Availability Form and submitting it together with their manuscript.

Upon submitting a revised manuscript, all authors should include ‘Data Availability Statement’ at the end of the final article file, which summarises whether and how these replication materials can be accessed.

Replication materials can be made available in the following ways:

• Via community resources (preferred), such as data and code repositories, providing a DOI or stable link to the replication materials in the Data Availability Statement. r3data.org is a registry of suitable data repositories that are: run by a legal entity (a sustainable institution such as a library or university); clarify access conditions to the user; and have long-term storage plans. Examples include: GitHub, Zenodo, Dataverse and the Open Science Framework. Many repositories allow the deposition of data with an embargo period before making the data publically available, if this is needed. A list of protected access repositories (where confidential data can be stored and managed) is available here.
• As supplementary material. Replication materials can be provided through the online submission system to be published as supplementary material, hosted by the journal. Note, however, that unlike data and code repositories, supplementary materials do not receive a DOI.  Files provided as supplementary material will not be copy edited or otherwise modified before publication.

Further points to consider:

• Replication materials should contain a “README” file, explaining what is included and how the code should be run, as well as a replication script that can be used to directly reproduce the results in the paper.
• Authors should also adhere to any additional requirements that may apply to them in the policies of their research funders and host institutions.
• AAS assumes no responsibility for the data uploaded to external repositories or hosted as supplementary material, as peer-review of these materials is not part of the editorial process. Authors are responsible for ensuring that the data are usable and suitably anonymised, the files uncorrupted, and for addressing any eventual questions from scholars wishing to replicate the authors’ results.
• Articles and supplementary material should not include sensitive information, for example personally identifiable data. De-identified data should be provided as far as that is safe and practical to do. If that is not possible, aggregated data derived from sensitive data should be provided. When anonymising data, be sure to take into account both direct and indirect identifiers.

Data Availability Statement

Upon revised submission, all authors should include ‘Data Availability Statement’ at the end of the final article file, which summarises whether and how these replication materials can be accessed.

Example statements are in Table 1 below, including those that give reasons why replication materials cannot be shared.

AAS and the Institute and Faculty of Actuaries (IFOA) assume no responsibility for the accuracy of the author’s Data Availability Statement, the resources that are linked to in the Data Availability Statement, or supplementary material provided by the author. Where the Data Availability Statement links to an external repository or another entity holding the resources, this does not imply an endorsement or warranty given by that organisation to the published article.

Table 1. Example Data Availability Statements.