Draft Concordat on Open Research Data

This response is submitted by the Royal Geographical Society (with IBG), the learned society and professional body for geography. The response is informed by a consultation with the Society’s research committee, the Editors of its journals, and members in higher education institutions across the UK.

We welcome the Concordat’s aim to establish voluntary principles which ‘help ensure that the research data gathered and generated by members of the UK research community is made openly accessible for use by others wherever possible in a manner consistent with relevant legal, ethical and regulatory frameworks and norms’. As we stress below, it is important that these are voluntary and not mandated.

We are pleased that the Concordat recognises some of the legitimate constraints on openness, and that appropriate degrees of openness will vary according to subject, disciplinary fields, and the type of data generated.

While we, and many within our discipline, welcome the aspirations set out in the Concordat, our community have also highlighted a number of challenges surrounding implementation that warrant further consideration.

 

General observations

• Not mandated: It is important that these principles and their implementation are and remain voluntary and not mandated.

• Greater recognition of diverse and plural types of data produced through different types of research practices: Although the Concordat acknowledges the diversity of research activities, recognition of the multiple types of data and research practices needs to be further embedded in the approach and the principles.

• Difficulty of delineating research as ‘publicly-funded’: This is not always straightforward. UK University researchers may be awarded grants that are, in turn, funded via multiple sources (for example, an institution’s membership fees, public funds, commercial sponsorship, or private donation). The extent to which research funded by these types of grants can be considered publicly funded is not always clear and needs to be recognised.

• Greater consideration of the role of expert judgement in decisions about making data open: This is relevant to all kinds of data, but particularly datasets linked to other qualitative analyses that could be important in the context of final data made available.

 

Observations relating to specific principles

• Resource and training implications of good data management (principle #2/ #9): Achieving an open research environment will require data management to be properly incorporated and appropriately resourced. Researchers will require additional support and training in data curation and management, the resource implications of this should not be underestimated. Universities and Learned Societies are well placed to advise on the most appropriate training for different types of data. Training costs could be included within research grant proposals and/or probationary training for early career researchers.

• Recognition of the labour (and associated resource implications) required to make open research data legible to external audiences (principle #3): The labour in making some sources of data (for example, a research diary) legible for external users may be considerable. Providing additional metadata and background material will also have workload implications. The Concordat does not address directly questions surrounding responsibility for resourcing open research data, and the processes that support it. Insufficient resource and improper data management are likely to be (and have been) major hurdles in making research data open.

• Data without metadata often have little value, better guidelines on compiling metadata (for different disciplines) will be essential for open data (principle #3): The provision of metadata must also be sensitive to disciplinary best practice and the type of data collected. For example, much research in human geography relies upon anonymity, which may become problematic in an open data environment where more metadata and background material is provided. This may particularly be the case with regard to data on topics such as activism, migration, racism, labour practices, environmental movements etc. Omitting such data may result in the exclusion of substantial amounts of interesting and important geographical research.

• Translation of research data between different international contexts (Principle #3): Greater attention could be paid to issues around the translation of research data between different international contexts; this would be a significant challenge for many researchers in geography.

• Reasonable first use: Applicability of a ‘short-well-defined period’ across disciplines (Principle #6): The right to ‘reasonable first use’ is only assured within a ‘short-well-defined period’, which may vary by subject and discipline. Depending on the type of research activity, it can take many years to collect data, analyse and publish it. The time frames involved may not be considered to be a very ‘short’ period. It is important that there are not unintended consequences where expectations of open data may discourage such research.

• Providing incentives to make datasets open (Principle #6): Researchers could be incentivised to make research data open, if their work is recognised by familiar metrics frameworks. For example, use of datasets by others could be tracked and recognised through DOIs, download figures, and possibly citations. In geography this already occurs in some fields, for example palaeoclimate, and datasets held by organisation such as the National Oceanic and Atmospheric Administration (NOAA) and Pangea.

• Better coordination is needed between funding bodies, journals and institutions about roles and responsibilities for storing and sharing data (Principal #8). Universities already deal with storing masses of data from research being undertaken, and generally comply with research funding bodies with respect to privacy, obligations etc. It is important there is not duplication of effort and resource.

• Formally valuing work done to achieve open data environments (Principle #10): The academic ‘credit system’ must come to recognise and value the work done, often by early career researchers, in developing these initiatives. This issue is currently underrepresented in the Concordat.