Overview of related University policies

The University of Bath recognises that research data are valuable assets and that good management of research data brings benefits to the researcher, the University, and the wider community. This Research Data Policy demonstrates the University’s support for the UKRI position that publicly funded research data are a public good that should be made openly available with as few restrictions as possible in a timely and responsible manner. The Policy implements the principles set out in both the Concordat on Open Research Data and the UKRI Common Principles on Data Policy and, as far as is possible, is aligned with the FAIR data principles.

The University recognises that it is the collective responsibility of researchers, doctoral and project supervisors, and the institution to ensure that research data are appropriately managed. The Library’s Research Data Service provides guidance and training on all aspects of research data management, and also manages the University’s Research Data Archive. Support can be obtained from the Research Data Team via research-data@bath.ac.uk. Digital, Data & Technology (DD&T) provide secure and resilient storage for research data, advice and support for data security, and facilities for advancing research computing.

The Policy and guidance should be read in conjunction with the University's Code of Good Practice in Research Integrity; the Intellectual Property Policy; the Institutional Code of Ethics; the IT Security Policy; the University's Data Protection guidance; guidance on Freedom of Information; and the Information Classification Framework.

The Policy guidance will be updated as new infrastructure to support research data management is developed.

Section 1

This policy applies to all research conducted by members of the University that is expected to result in published findings of lasting scholarly value.

1.1. Research is defined according to the internationally accepted OECD Frascati Manual 2015 as "Creative and systematic work undertaken in order to increase the stock of knowledge, including knowledge of humankind, culture and society, and to devise new applications of available knowledge." Research must be “novel, creative, uncertain, systematic, transferable, and reproducible”.

1.2. Research data are defined by the Concordat on Open Research Data as “the evidence that underpins the answer to the research question, and can be used to validate findings regardless of its form (e.g. print, digital, or physical).” These data may be quantitative information, software, or qualitative statements collected by researchers in the course of their work by experimentation, observation, modelling, interview, or other methods, or information derived from existing evidence.

1.3. This policy will only apply to research, as defined above, conducted by undergraduate or taught postgraduate students when the findings of the research are published in peer-reviewed publications. Project supervisors are encouraged to discuss the potential publication of supporting research data with students as part of project planning.

1.4. The policy does not normally apply to research conducted by the Students’ Union or Professional Services, such as surveys used to improve service provision, except where findings are published in a peer reviewed journal.

1.5. The policy applies to all current and future research, and replaces the previous Research Data Policy, which was approved in April 2014. Consideration will also be given to legacy research outputs that have value to the University.

1.6. The policy does not apply to consultancy services conducted for third party organisations including those carried out using or by University Core Research Facilities.

Section 2

The University of Bath Principal Investigator or Doctoral/Project Supervisor associated with a research project is the Data Steward for that project and is ultimately responsible for research data management.

2.1. The Data Steward can delegate specific responsibilities and this should be outlined in the project’s Data Management Plan.

2.2. In practice, the day-to-day management of research data is likely to be a shared responsibility, with all researchers having a personal responsibility for the data they create or use. In the case of students, these practical responsibilities should be overseen by their doctoral or project supervisor.

2.3. In the absence of the Data Steward, responsibility for stewardship of research data, including arbitration on the retention of legacy data held on University systems, will transfer to the Head of Department or Dean of School.

2.4. Where research is conducted in collaboration with external research partners, or where there are multiple co-Investigators at the University, a single Data Steward at the University of Bath must still be assigned. This Data Steward is responsible for the management of research data that is under the University’s control.

2.5. Instrument Specialists in University Core Research Facilities will be responsible for the systems on which research data are created, but the Data Steward for the project will be responsible for the management of the data created by them, or by members of their research group, in accordance with this policy. This may involve the Data Steward transferring a copy of data onto resilient project storage prior to the end of the project, and ensuring the long-term preservation and publication of data via an appropriate data archive.

2.6. Before leaving the University at the end of a contract or completion of studies, researchers must first discuss management and retention of all research data for which they are responsible with their line manager, doctoral or project supervisor, or Head of Department as appropriate, to determine whether a copy of the data will be retained by the University for continued study, and whether the researcher will take a copy. These decisions must conform with requirements and restrictions arising from applicable legislation, funding, contracts, ethical approval, and policies.

Section 3

Researchers are responsible for making themselves familiar with and adhering to legislation, regulatory requirements, contractual obligations, ethical approvals, and funder policies that pertain to their research data. Where research involves the use of data owned by a third party, researchers must also abide by any applicable licences or terms of use governing the data.

3.1. Where multiple requirements apply to a particular research project, legislative requirements, such as Data Protection, supersede all policy requirements.

3.2. New contractual agreements for collaborative projects and studentship agreements with external bodies must comply with funder and University policies on research data.

3.3. Researchers in need of advice on policies and requirements should contact the Library’s Research Data Service, IT Security, the Data Protection team, the Freedom of Information Officer, their Departmental Research Ethics Officer, or Research and Commercialisation Contracts as appropriate.

Section 4

Data Management Plans must be written for all research projects that are expected to process data, irrespective of whether submission of such plans is required when applying for research funding.

4.1. Reference to Data Management Plans in this policy can equally refer to Output Management Plans or Software Management Plans, where these are either specifically required by a funder or are more applicable to the research respectively. Similarly, research funders may also refer to such plans by a variety of names, e.g. data management and sharing plan.

4.2. Data Management Plans must explicitly address the creation, management, confidentiality, retention, and publication of data, software, and, where applicable, other valuable outputs, created or collected during a project. Plans should also set out responsibilities for the management and control of data.

4.3. If a brief Data Management Plan is included in a grant application, this should be expanded once funding is awarded, particularly for EPSRC funded research. For collaborative projects involving one or more external partners, access to and publication of research data should normally be included in collaboration agreements, and data sharing agreements should be entered into where required.

4.4. Data Management Plans should consider the approach to data preservation and publication at the project outset or grant proposal stage, in order to mitigate any potential barriers to data sharing. Justifiable reasons for restricting access to retained datasets, or extended periods of exclusive use prior to data publication, should be set out in the Data Management Plan.

4.5. Where applicable, funder- or doctoral-specific templates and guidance for creating a Data Management Plan should be used. Should no specific template exist, the University of Bath provides generic templates.

4.6. All members of a research group must familiarise themselves with and adhere to the Data Management Plans covering their research.

4.7. Data Stewards must review Data Management Plans at least annually during the life of the project to ensure adherence and continued relevance to the research.

Section 5

Where possible researchers should seek to recover the direct costs of managing research data from the research funder.

5.1. The justified costs associated with delivering the Data Management Plan should be considered at the earliest opportunity.

5.2. The University provides a limited amount of active and preservation storage for free to Data Stewards, but charges may be incurred for requirements exceeding those limits (currently 1TB).

5.3. Legitimate costs associated with managing and publishing data include:

5.3.1. storage of active data in volumes that exceed institutional provision,

5.3.2. data publication, where there is a charge associated with deposit in a data archive, or where volumes exceed institutional provision,

5.3.3. transcription of interviews,

5.3.4. recruitment of specialist staff to support the management of complex datasets and the preparation of these data for archive and publication.

Section 6

Researchers must ensure that active research data are stored securely and protected from loss and unauthorised access. Data Stewards should ensure that access is not limited to a single person.

6.1. Research data should be stored appropriately for its level of classification in line with the Information Classification Scheme.

6.2. To minimise the risk of data loss, and to maintain the integrity and authenticity of data, digital material must be protected through the use of resilient electronic systems, such as University managed file storage. Removable media should only be used for temporary storage and transportation of data. Appropriate security measures, such as encryption, should be used to prevent unauthorised access to confidential and personal data. Where possible, non-digital material should be digitised at the earliest possible opportunity.

6.3. Researchers must ensure that research data are stored in a jurisdiction that offers the same levels of protection that are available in the UK. For example, any research data containing personal data must not be transferred to any jurisdiction outside the European Union unless there are adequate safeguards in place, or without explicit informed consent to do so from the individuals the data concerns. If there is a need to transfer personal data outside the European Union in other circumstances, or if there is any uncertainty, researchers must first contact the Data Protection Team for advice.

6.4. Researchers must not try to avoid Freedom of Information requests by removing research data from the UK.

6.5. Where external security requirements demand more stringent access restrictions as a condition of using a dataset, Data Stewards should contact the IT Security Team or the Research Data Service prior to accepting the dataset.

Section 7

Researchers should provide sufficient explanatory documentation and descriptive metadata about their research data to ensure that data are understandable, re-usable, and, where appropriate, discoverable, by peer researchers.

7.1. Documentation and descriptive information should be recorded as early as possible during data creation.

7.2. Relevant metadata standards or minimal information sets should be used where applicable.

7.3. Documentation should include details of how the research data were collected and processed, and should enable peer researchers to understand the data sufficiently to be able to repeat or extend the research, without consulting the original data creators.

7.4. Consistent and meaningful file naming conventions should be used, adhering to any conventions already agreed by collaborating researchers or research Facilities.

Section 8

Research data must usually be retained after project completion if they substantiate research findings, are of acknowledged long-term value, support a patent application, or if other legal or regulatory requirements for data retention apply.

8.1. There may be legitimate reasons why data falling into the above category must instead be destroyed, such as licence restrictions on the use of third party data or commitments made to research participants. Where there are legal or regulatory requirements for data destruction, these override data retention policies.

8.2. Retained data supporting publications should be sufficient to enable peer researchers to validate the research findings.

8.3. Supervisors of doctoral or project students must agree, prior to graduation, which of their students’ data will be retained by the University. Data retention must comply with any applicable studentship agreements, funder policies, and legal, regulatory, contractual, intellectual property, or other obligations.

8.4. The Research Data Service provides guidance on how to determine the potential long-term value of a dataset. Data with long-term value might include unique or unreproducible data, data expected to be re-used regularly, new data contributing to an existing collection, or data representing a landmark discovery or precedent.

8.5. Where research data are not retained, methods of data destruction should be appropriate for the security classification of the data in line with the Information Classification Scheme.

Section 9

Unless legal, regulatory, or funder requirements specify otherwise, data selected for retention must be preserved for 10 years from the end of a project or the publication date of any research findings based upon them, whichever is the later, after which retention will be reviewed.

9.1. Retained data should be submitted to an appropriate data archive, data repository, or database for long-term preservation. That archive, repository, or database will then be responsible for managing long-term data preservation.

9.2. Recommended time points for the upload of retained data to a data archive, data repository, or database include completion of a work package, submission of an article or other output for peer review, conclusion of research funding, or submission of a PhD Thesis. Researchers should liaise with the archive, repository, or database administrators on managing the timing and openness of publication of uploaded data, in accordance with Section 10 of this policy.

9.3. The use of project websites, personal computing equipment, University Research Facilities, or University-managed active research storage (e.g. the X:drive) for long-term data retention should be avoided.

9.4. As far as practicable, retained research data should be converted to open, lossless, or non-proprietary formats for preservation to facilitate long-term usability. Where this is not possible or reasonable practicable, details of any proprietary software required to process the data must also be retained, although it is not normally necessary to also retain or provide access to the software.

9.5. The Library’s Research Data Service will be responsible for the preservation of research data deposited in the University’s Research Data Archive, and will make decisions on the long-term retention of data held in the Archive.

9.6. Examples of external requirements that would supersede the usual 10 year preservation period include, but are not limited to, some NHS requirements to either destroy research data after 5 years or retain research data for 15 years; MRC requirements to retain some data types for 20 years; or specific commitments made to research participants.

Section 10

Research data that have been selected for retention must be made openly available with as few restrictions as reasonable practicable, unless this would breach legislative, regulatory, contractual, ethical, or other obligations, or unless the cost of doing so would be prohibitive.

10.1. Research data selected for retention must, where possible, be made openly available from the first publication of research findings based on the data or within 12 months of project completion, whichever is the sooner. Funder policies stipulating more rapid data publication, or where data relate to public health emergencies, override this requirement.

10.2. Research data selected for retention should be published via an appropriate data repository in a citable format, and must be issued with a persistent identifier such as a Digital Object Identifier or accession number.

10.3. Justifiable reasons for restricting access to data include, but are not restricted to, maintaining confidentiality, protection of individual privacy, commercial sensitivity, licence restrictions on third party data, protection of sensitive environmental or cultural sites, protection of vulnerable groups or species, and disproportionate costs.

10.4. Application of access restrictions to all files in a dataset should be avoided where possible. Appropriate anonymisation and redaction strategies, or separation of restricted and open files, should be used to maximise open access to published data where practicable.

10.5. Where data has a commercial value or supports a commercial output such as a patent then public disclosure of the data may be delayed. The Research Commercialisation Managers in Research & Innovation Services can advise on the commercial value of research data.

10.6. Researchers should acknowledge any Facilities that they used when publishing outputs based on data created in these Facilities.

10.7. Publication of research data should normally be considered in all new consortium and collaboration agreements, and when consent is sought from research participants.

10.8. The Library’s Research Data Service can assist with identification of a suitable location for data publication and can provide advice on the available exemptions that may be used to justify withholding of research data.

Section 11

A data access statement describing how and on what terms any supporting research data may be accessed must be included in associated publications based upon supporting data.

11.1. Where available a persistent identifier for a dataset, such as a Digital Object Identifier or accession number, must be included in this statement.

11.2. Guidance on data access statements, including examples for different types of access restriction, is available from the Research Data Service web pages.

11.3. Where research includes the re-use of third party data, the data access statement should link either directly to the published dataset(s), or to documentation describing how a peer researcher could gain access to the same dataset(s).

11.4. All sources of data that support a journal article’s findings should be included as formal citations in the list of references for the article, including those described in the data access statement.

Section 12

Published research data should be licensed to permit the widest possible reuse, whilst ensuring that original data creators receive appropriate credit.

12.1. Licences granted should have as few restrictions as possible and must specify that, at a minimum, users of the data must acknowledge the originator of the data in any publication or derived work. This will facilitate future tracking of the citation, re-use, and impact of data.

12.2. Except where it is a condition of a grant or contract, exclusive rights to the ownership or use of research data must not be assigned, licensed, or otherwise transferred to external parties such as sponsors, publishers, collaborators, or external data archives.

Section 13

Published research data should be registered with the University, and if appropriate reported to Researchfish, even if access to the data is restricted.

13.1. Data published via the University’s Research Data Archive are automatically registered with the University. Externally published data should be registered with the University within 12 months of publication of the dataset.

13.2. Registration involves providing the University, via a dataset record in Pure, with information about what the data are, who created and is responsible for them, and where the data are preserved and/or published. This record should include a link to the published data.

13.3. Registration of externally published research data allows the Library’s Research Data Service to ensure that these University assets remain preserved in the rare event that a data archive ceases to operate, and to monitor the wider impact of data publication.

13.4. Researchers are expected to report to Researchfish all published datasets arising from applicable projects, under ‘Research Databases & Models’.

Owner: Library Research Data Service Version number: 2 Approval Date: 6 February 2019 (originally approved April 2014) Approved By: Senate Date of last review: February 2019

The wording of this research data policy guidance is licensed under a Creative Commons Attribution 4.0 International Licence. All logos, page design, and other intellectual property remain under copyright to the University of Bath.