Recommendations for Sharing Network Data and Materials

"Researchers should share the network data and materials necessary to reproduce reported results via a publicly accessible repository when an associated manuscript is published" (p. 404).

I endorse this recommendation

Neal, Z. P., Almquist, Z. W., Bagrow, J., Clauset, A., Diesner, J., Lazega, E., Lovato, J., Moody, J., Peixoto, T. P., Steinert-Threlkeld, Z., and Teixeira, A. S. (2024). Recommendations for sharing network data and materials. Network Science, 12, 404-417. https://doi.org/10.1017/nws.2024.16

The first page of the article titled "Recommendations for sharing network data and materials"

Background

In 2023, the International Network for Social Network Analysis (INSNA) requested that Zachary Neal form a working group to develop recommendations for sharing network data. Members of the working group were recruited from both the "social network analysis" (SNA) and "network science" (NetSci) traditions to ensure broad intellectual, demographic, and geographic representation.

Throughout 2023, the working group met virtually to draft a preliminary set of recommendations. The development of recommendations was guided in part by the TOP guidelines, the FAIR principles, and the CARE principles. The group also considered the current data sharing expectations of journals that commonly publish network research.

The preliminary recommendations were submitted on 15 December 2023 to Network Science for review by the editorial board, with the review process overseen by Associate Editor and INSNA President Dr. Laura Koehly. The working group received feedback from Dr. Koehly and three anonymous editorial board members in April 2024, and submitted a revised set of recommendations in May 2024. The working group received additional feedback from the same three anonymous editorial board members in July 2024, and submitted a revised set of recommendations, which were accepted for publication in September 2024.

Detailed Recommendations

We offer four related recommendations for researchers:

Researchers should share their network data and materials, but may restrict access when necessary to prevent harm, comply with regulations, or protect privacy.
Researchers should share the network data and materials necessary to reproduce reported results, but researchers may choose to share additional data and materials that facilitate their re-use for other purposes.
Researchers should share their network data and materials when an associated manuscript is published, but researchers may choose to share earlier in the dissemination and peer review process.
Researchers should share their network data and materials in a repository that is publicly accessible, searchable, versionable, and offers Digital Object Identifiers (DOIs).

We also offer recommendations for key institutions to support researchers:

Journals: Require authors to follow these guidelines, but exercise discretion in enforcing this requirement
Universities: Provide assistance with sharing data, and reward data sharing as a form of research productivity.
Associations: Encourage members to share data, and reward sharing as a form of service.

Rationale

Beyond the frequently-noted benefits of transparency and openness, we believe that the routine sharing of network data offers a "CURE" for restrictions on data that restrict the advancement of network science:

Sharing of network data and materials will ensure that network research is in compliance with increasingly common mandates from public funding agencies that support network research.
Sharing network data and materials associated with a published study facilitates readers’ understanding of the study by allowing readers to find details not typically contained in a manuscript and allowing students to learn how to use the methods.
Shared network data and materials are essential for verifying the reproducibility of a study’s findings, which is the bedrock of extending research in new or deeper directions.
Sharing network data and materials promotes the efficiency of network research by reducing the effort necessary for future studies, while also facilitating the formation of collaborative teams working on common data or with common materials.

Concerns

In developing these recommendations, the working group acknowledged that there are some reasonable concerns about sharing data:

Time: Preparing data for sharing can involve a significant amount of time. However, this step is as important as the other steps in the research process (e.g., data collection, analysis), and should be treated as a routine part of the research process.
Data: Collecting network data can require significant effort, leading researchers to be reluctant to simply "give away" their data. However, these recommendations do not ask researchers to give away their data. We only recommend that researchers share the specific pieces of data used in a given publication, and not necessarily an entire dataset. Additionally, when data are shared, they should be shared with a license that specifies how they can be used.
Privacy: Network data are information-rich, and risks of re-identification may lead researchers to be reluctant to share even de-identified data. The risk of network data revealing private information must be balanced against the benefits of research transparency, and such trade-offs must be considered on a case-by-case basis.
Trust: Recommending that researchers share their data may be misinterpreted as an expectation of suspicion of fraud, or a lack of trust in fellow scientists. However, rather than as a practice for preventing bad academic behaviors, data sharing should be viewed as a practice for promoting good academic behaviors of transparency and collaboration.

Institutional Endorsements

Individual Endorsements

To endorse these recommendations and be added to this list, please send an email to zpneal@msu.edu from your institutional email account.

Zachary P. Neal, Michigan State University
Zack W. Almquist, University of Washington
James Bagrow, University of Vermont
Aaron Clauset, University of Colorado
Jana Diesner, Technical University of Munich
Emmanuel Lazega, Sciences Po
Juniper Lovato, University of Vermont
James Moody, Duke University
Tiago P. Peixoto, Interdisciplinary Transformation University
Zachary Steinert-Threlkeld, University of California - Los Angeles
Andreia Sofia Teixeira, Northeastern University London

Recommendations for Sharing Network Data and Materials

"Researchers should share the network data and materials necessary to reproduce reported results via a publicly accessible repository when an associated manuscript is published" (p. 404).

Neal, Z. P., Almquist, Z. W., Bagrow, J., Clauset, A., Diesner, J., Lazega, E., Lovato, J., Moody, J., Peixoto, T. P., Steinert-Threlkeld, Z., and Teixeira, A. S. (2024). Recommendations for sharing network data and materials. Network Science, 12, 404-417. https://doi.org/10.1017/nws.2024.16

Background

Detailed Recommendations

We offer four related recommendations for researchers:

Researchers should share their network data and materials, but may restrict access when necessary to prevent harm, comply with regulations, or protect privacy.

Researchers should share the network data and materials necessary to reproduce reported results, but researchers may choose to share additional data and materials that facilitate their re-use for other purposes.

Researchers should share their network data and materials when an associated manuscript is published, but researchers may choose to share earlier in the dissemination and peer review process.

Researchers should share their network data and materials in a repository that is publicly accessible, searchable, versionable, and offers Digital Object Identifiers (DOIs).

We also offer recommendations for key institutions to support researchers:

Journals: Require authors to follow these guidelines, but exercise discretion in enforcing this requirement

Universities: Provide assistance with sharing data, and reward data sharing as a form of research productivity.

Associations: Encourage members to share data, and reward sharing as a form of service.

Rationale

Beyond the frequently-noted benefits of transparency and openness, we believe that the routine sharing of network data offers a "CURE" for restrictions on data that restrict the advancement of network science:

Sharing of network data and materials will ensure that network research is in compliance with increasingly common mandates from public funding agencies that support network research.

Sharing network data and materials associated with a published study facilitates readers’ understanding of the study by allowing readers to find details not typically contained in a manuscript and allowing students to learn how to use the methods.

Shared network data and materials are essential for verifying the reproducibility of a study’s findings, which is the bedrock of extending research in new or deeper directions.

Sharing network data and materials promotes the efficiency of network research by reducing the effort necessary for future studies, while also facilitating the formation of collaborative teams working on common data or with common materials.

Concerns

In developing these recommendations, the working group acknowledged that there are some reasonable concerns about sharing data:

​

Time: Preparing data for sharing can involve a significant amount of time. However, this step is as important as the other steps in the research process (e.g., data collection, analysis), and should be treated as a routine part of the research process.

Institutional Endorsements

Individual Endorsements