Data Management Plan (DMP) - Guidelines for researchers
Managing and sharing research data as openly as possible is one of the principles of good scientific practice. The SNSF adheres to this principle, as stated in Article 47 of its Funding Regulations (PDF): in stating that"[…] grantees are obliged to make available to the public in an appropriate manner the research results obtained with the help of SNSF funding, […]". The SNSF has set out the criteria it expects funded researchers to meet in its Open Research Data Policy statement For the implementation of these principles, the SNSF favours a bottom-up approach. It provides best practice guidelines and gives each scientific community sufficient flexibility in defining and applying its own standards. In particular, the best way of managing and sharing data depends on the research field.
The aim of a Data Management Plan (DMP) is to plan the life cycle of data. It offers a long-term perspective by outlining how data will be generated, collected, documented, shared and preserved. The SNSF provides a template to help researchers complete their data management plan. Each project's DMP will refer to discipline specific standards and practices and thus its content may be different.
All data collected and generated during the course of the research on which publications are based must be shared, provided no legal, ethical, intellectual property or other clauses prevent this. This data should be made available as soon as possible, at the latest together with the relevant scientific publication, and should be archived in recognized scientific data repositories that meet the FAIR data principles (IR 11.8). Data can be raw or processed, depending on the project and the discipline. Datasets must always be carefully documented with associated metadata, such that other researchers understand how the data was collected, as well as under which conditions and how it can be re-used. If specific tools are needed to re-use the data, this needs to be documented and, if possible, the tools made available. In any case, the provided data and documentation (metadata, code, etc.) must be sufficient to ensure their reusability. Researchers are asked to explain in their DMP wherever these requirements cannot be met.
Data sharing – best practices
To facilitate the discovery, access, re-use and citation of datasets, it is important that the publication of research data follows a set of clearly defined and broadly applicable best practices. The FAIR Data Principles define a range of qualities a published dataset should have in order to be Findable, Accessible, Interoperable and Reusable (see Explanation of the FAIR Data Principles (PDF)). The SNSF expects researchers to share their data according to the FAIR Data Principles on publicly accessible, digital repositories. It is important to note that the FAIR Data Principles do not require researchers to share all their data without any restrictions. Rather they advocate applying a standard procedure when sharing research data for reuse, so that humans and computer systems can easily find, interpret and use them under clearly defined conditions. The FAIR Data Principles were adopted by a growing number of research funding organisations (e.g. Horizon Europe, NIH, FWF, Swedish Research Council).
2. The SNSF's Data Management Plan
With its Open Research Data (ORD) policy, the SNSF requires a data management plan (DMP) for approved applications. It should provide information on the data life cycle. In order to account for different data management practices between disciplines, the SNSF has defined minimum standards for the structure and content of the information to be provided.
2.1. How to submit a DMP
For funded applications, grantees receive an invitation to submit their DMP on the mySNF platform. The DMP has to be written in the same language as the research plan. Submitting a DMP is a condition for the release of funds.
The DMP form comprises four sections: (1) data collection and documentation, (2) ethics, legal and security issues, (3) data storage and preservation, and (4) data sharing and reuse. Sub-questions and online help texts will help researchers to complete the form (see details (PDF)). The “questions you might want to consider” will help you to complete the form. Grantees must enter a DMP that is plausible, suits their project and meets the standards set by their research community. Eventual missing or inaccurate statements need to be added/revised at the latest before the release of the payment of the second instalment. A definitive and updated version of the DMP must be provided by the end of the project grant.
Some research projects do not produce or reuse any data. If this is the case, grantees do not have to complete the whole DMP form. However, they are asked to explain why they do not expect to generate or reuse any data in their proposed research.
Some data cannot be shared because grantees are bound by legal, ethical, copyright, confidentiality or other clauses. They will be asked to explain their specific constraints.
If the grant is funded under (i) the Lead Agency or Weave process to a foreign partner organization which serves as Lead Agency, (ii) as part of an ERA-NET call or (iii) as part of a European Partnership, the DMP has to cover the research data, which are collected, observed, generated or reused in the Swiss part of the project and has to comply with the SNSF Open Research Data Policy.
2.2. Assessment of the DMP
The DMP is a requirement for the release of funds and is assessed by the SNSF administrative offices. Eventual shortcomings in the DMP have to be addressed at the latest before the release of the payment of the second instalment. In such a case, a task to revise the corresponding specific sections of the DMP is sent to the grantees on the SNSF platform.
2.3. Lifetime management
The DMP remains editable during the entire lifetime of the grant. Its contents can be adapted as the project evolves.
In any case, researchers will be prompted to update their DMP at the end of the grant. In this final version, the SNSF expects that the management of the data, which were collected, generated and observed during the course of the project, is described conclusively. This updated version will be assessed together with the final scientific report. The SNSF Administrative Offices retain the right to request additional information and/or amendments to the contents of the final DMP.
3. Examples of data management plans
4. Eligible Costs
The SNSF is aware that sharing research data requires resources. Therefore, applicants can request at the time of submission funding for the preparation of research data in view of its archiving, and to the archiving itself, in data repositories complying with the FAIR data principles and that do not serve any commercial purposes (IR 2.13). However, the SNSF contributes to these archiving costs only on the basis of a one-time payment at the time of data upload on the repository. The SNSF does not cover any subsequent costs. The SNSF may allocate up to CHF 10,000 for these activities. The costs must be taken into account at the time of submission of the application. They cannot be covered by a supplementary grant.
5. Examples of repositories that comply with the FAIR Data Principles and are non-commercial
Requesting that researchers apply the FAIR Data Principles in every detail is an ambitious policy. In addition, finding the "perfect" repository providing all necessary features to host FAIR data can be a challenge. To make the transition towards FAIR research data easier, the SNSF decided to define a set of minimum criteria that repositories have to fulfil to conform with the FAIR Data Principles (see checklists below).
Four repositories which accept datasets from different research fields and fulfill the SNSF requirements are shown here (PDF). It is, of course, possible to archive data on other (field-specific) repositories. Researchers can proceed as follows to ensure that the chosen repository is in line with the SNSF requirements.
5.1. Checklist to identify repositories complying with the FAIR Data Principles
- Are datasets (or ideally single files in a dataset) given globally unique and persistent identifiers (e.g. DOI)?
- Does the repository allow the upload of intrinsic (e.g. author's name, content of dataset, associated publication, etc.) and submitter-defined (e.g. definition of variable names, etc.) metadata?
- Is it clear under which licence (e.g. CC0, CC BY, etc.) the data will be available, or can the user upload/choose a licence?
- Are the citation information and metadata always (even in the case of datasets with restricted access) publicly accessible?
- Does the repository provide a submission form requesting intrinsic metadata in a specific format (to ensure machine readability/interoperability)?
- Does the repository have a long-term preservation plan for the archived data?
5.2. Checklist to identify non-commercial repositories
The first step is to consult www.re3data.org, where most repositories are listed.
- Under the tab "Institutions", check if a commercial entity is involved in 'general' or 'technical' responsibility (categories "Type of institution" and "Type(s) of responsibility")
- If not, SNSF considers the repository to be non-commercial (even if ‘funding’ or ‘sponsoring’ is provided by a commercial entity).
- If yes, the SNSF considers the solution to be a commercial repository (see details (PDF))
If the repository is not listed on www.re3data.org, the repository should be contacted to clarify this point. Researchers should also suggest that the repository be included in www.re3data.org.