This article is the final update in our #RoadToOpen series for 2024. Our commitment to the journey of becoming a fully open access publisher is ongoing and the series will continue into 2025, sharing our learning as we respond to the developing research environment. This update builds on the successful transition of our maths portfolio to open access under the Subscribe to Open (S2O) model, with increasing attention now on supporting open science.

Open Science Indicators in the mathematics portfolio

In December 2022, open access publisher PLOS launched “Open Science Indicators” (OSI), a dataset by which changes in the availability of data, code, and preprints associated with research articles can be measured over time.

Inspired by the PLOS dataset, EDP Sciences began tracking key open science data points for the 5 mathematics journals we co-publish with the Society of Applied and Industrial Mathematics (SMAI): Mathematical Modelling of Natural Phenomena, ESAIM:Control, Optimisation and Calculus of Variations, ESAIM:Mathematical Modelling and Numerical Analysis, ESAIM:Probability and Statistics, RAIRO-Operations Research.

In March 2023, with the approval of the Editors-in-Chief, we put in place a new data and code sharing policy for the 5 mathematics journals, encouraging the sharing of data and code in trusted public repositories. If the authors have deposited their data/code in a public repository, and declared it at the time of submission, then it is mandatory to include a Data Availability Statement and a reference to the data/code in the reference list of the article (“data citation”). 

We implemented a process to track articles whose authors declared at the time of submission that they have deposited data/code in a public repository. During a pilot phase in 2023-2024 all articles flagged accordingly were manually checked by the publisher before publication, and article proofs inconsistent with the data policy were sent back to the author for correction (e.g. missing Data Availability Statement even though the author reported sharing data/code, missing data citation, missing DOI or persistent identifier in the statement, data deposited in a personal/institutional repository rather than in a public repository, broken link, link pointing to other objects than data/code, etc.),.  

Analysing uptake of new policies using the Open Science Indicators

Inspired by PLOS, we are now reporting for the first time:

• the deposit of data and code in a repository by authors of the 5 mathematics journals since 2023.
• the peer-reviewed preprints in Peer Community In that are eventually published in the"PCI-friendly” journal Mathematical Modelling of Natural Phenomena.

Note: unlike PLOS we do not track preprints deposited in preprint repositories. In mathematics, there is a high prevalence of preprint sharing in repositories such as arXiv.

The figure below shows the evolution of the number of articles reporting data and code sharing in 2023 and 2024 in the 5 SMAI/EDPS journals MMNP, ESAIM:COCV, ESAIM:M2AN, EASIM:PS, RAIRO-RO. The specific repositories are indicated. 

In 2023, 15 articles published in the 5 journals reported data and/or code sharing, while in 2024, 26 articles (articles published or in press by December 2024) reported data and/or code sharing. The increase is due primarily to the implementation of the new policy in March 2023, and the lag until articles with data statements were subsequently published. The practices of the research community, too, are ever evolving, and it is reasonable to assume that there has been an increase in data/code sharing and declaration practices generally.

Encouraging the adoption of open science practices

The number of authors depositing data/code in a repository and declaring these in their article represents a small fraction of the total number of articles published in the EDP Sciences/ SMAI maths journals, only approximately 3% in 2023, and 5% in 2024. In comparison, 6 years after the launch of the OSI program, PLOS reported approximately 30% data sharing, and 15% code sharing, as a proportion of all published articles.

It is the work of open science leaders to first implement these sharing policies, and then encourage their adoption. Starting in 2025, EDP Sciences and SMAI will report on these indicators in each Subscribe To Open Transparency Report, charting this evolution and encouraging the deposit of data and code as a habitual part of the publishing process. (Read the 2024 S2O Transparency Report here)

Mathematics is a discipline that has long championed open access, with preprint sharing in repositories already ubiquitous. The deposit of data and code is expected to become normalised, promoting inter- and multi-disciplinary collaboration and scientific advancement. In the future, we hope to further automate our OSI workflows, and extend this dataset to other journals published by EDP Sciences.