Nelson Data Management 12

[Tyler Bannowsky]

Federalagencies should take coordinated action to ensure that data sharing policies created in response to the 2022 Nelson memo incentivize high-quality data management and sharing plans (DMSPs), include robust enforcement mechanisms, and implement best practices in supporting a more innovative and credible research culture.

Without independent research, we do not know if the AI systems that are being deployed today are safe or if they pose widespread risks that have yet to be discovered, including risks to U.S. national security.

We want to inform the University of Arizona community about the impacts of the new memorandum from the Office of Science and Technology Policy (OSTP) that was published in August 2022. The policy directs US federal agencies to develop plans for how they will ensure free, immediate, and equitable access to federally funded research.

The 2013 Holdren Memo required federally funded researchers to make their research papers publicly available no later than 12 months after publication. This policy applied to agencies with research and development (R&D) budgets of $100 million or more and required researchers to develop data management plans.

The 2022 OSTP/Nelson Memo is now directing federal agencies with any level of R&D budget to create policies requiring that funded researchers provide immediate public access to research papers and data upon publication.

A working group with members from Research Innovation and Impact (RII), UA Libraries (UAL), the Associate Deans for Research Council, University of Arizona Health Sciences (UAHS) Research Administration, the Data Science Institute, and others is monitoring how federal agencies, philanthropic societies, libraries, the broader academy, publishers, and scholarly communications organizations and advocacy groups are responding to the policy requirements. The working group will use this page to share news and resources as appropriate for the UA community.

Open access provides both free and unrestricted online access to read published research outputs with the addition of reduced copyright restrictions. Open access publications are released with a copyright license (such as a Creative Commons license) that allows some type of reuse, re-distribution, modification, or adaptation.

Scientific data is defined as: "the recorded factual material commonly accepted in the scientific community as of sufficient quality to validate and replicate research findings. Such scientific data do not include laboratory notebooks, preliminary analyses, case report forms, drafts of scientific papers, plans for future research, peer-reviews, communications with colleagues, or physical objects and materials, such as laboratory specimens, artifacts, or field notes."

Research data management has emerged as a need among academic researchers and librarians are building skills in response. This one-day workshop will assist liaisons to identify their existing skills and mindsets that transfer to research data management services and then create a learning plan for the RDM specific knowledge needed to serve their subject disciplines. Contact ACRL Program Officer Chase Ollis at col...@ala.org or 800/545-2433 ext. 2521 to discuss dates and locations, pricing, and complete workshop details.

Engaging Liaison Librarians: Identifying Impact of an Research Data Management Educational Intervention. By Megan Sapp Nelson and Abigail Goben in International Federation of Library Associations. Kuala Lumpur. August 2018.

journal articles and academic publishing models, the changes to federal data sharing policies outlined in the Nelson Memo have been comparatively underexplored. Statements from major organizations such as the AAU or AAAS, for example, mention data either in passing or not at all. Yet the new data policies will meaningfully affect researchers, have significant implications for the business model of generalist repositories, and may provide hope for domain repositories, many of which currently struggle to find financial sustainability as the grant-funding that created them dries up.

Current federal data sharing policies are based on the 2013 Holdren Memo. The Holdren Memo is broadly similar in its aims to the Nelson Memo, but has several key differences both in overall scope and in the language of specific sections.

Each of these concerns point to the essential role that research cultures play in data sharing. Indeed, even as the technical infrastructure to enable data deposit and reuse has matured, the cultural infrastructure that encourages researchers to change their priorities and practices have lagged behind. This is true even in fields where funders require publicly accessible data deposit, in part because of the challenges funders face in enforcing compliance. Absent stronger enforcement or substantive transformation in research cultures, or in all likelihood, sustained investment in both, data sharing may remain spotty. Neither topic is mentioned in the Nelson Memo, so it appears that the policies of individual agencies will be important factors in whether its most transformative possibilities become reality.

Domain and general repositories are an essential and now familiar part of the data sharing infrastructure. However, as our colleagues Roger Schonfeld and Rebecca Springer have pointed out, even the largest generalist repositories have struggled to develop sustainable business models. Specialized domain repositories often rely on grants for start-up funds. Even well-established domain repositories face financial challenges as those funds run dry. Though smaller and thus cheaper to maintain than generalist repositories, they also appeal to a narrower user base and have little to no full-time staff to make needed updates to the platform. The Nelson Memo specifies that federal grantees should be allowed to include costs associated with submitting, curating, managing, and publishing data into their research budgets, so both domain and generalist repositories could see opportunities for revenue growth.

Current policies at major agencies such as NSF and NIH encourage researchers to deposit in domain repositories when available, while leaving the ultimate choice of repository to grantees. While it is possible that some agencies may decide to create their own repositories, the Nelson Memo signals that current approach will likely continue. In theory, this should benefit domain repositories, which may have opportunities to raise revenue for deposit and even develop fee-based services. To maximize the possibilities of the Nelson Memo, domain repositories will need to convince researchers to designate them as the location of deposit in data management plans and grant budgets. Many domain repositories do not currently have well-developed marketing and communications strategies, which could hamper their ability to do the kind of outreach that will be required to maximize these opportunities. Moreover, it does not appear that federal agencies are expected to provide additional funds to costs associated with data sharing: those expenses will likely eat into research awards. Domain repositories will need to assess whether they can price deposit and services competitively, which could be a challenge given the economies of scale that generalist repositories can leverage.

It is quite possible that generalist repositories will emerge as the main beneficiaries of changes to federal policy. Some evidence suggests that researchers tend to prefer institutional and generalist repositories, and depending on how the issue of data that is not associated with publication plays out, a large percentage of the data newly subjected to data sharing requirements could come from fields without existing domain repositories. Both commercial generalist repositories like Mendeley Data or Figshare, and not-for-profit options such as ICPSR, Dryad, or Zenodo have capacity to adopt new features, accept a wide range of data types, and integrate data deposit into publication workflows.

The full effects of the Nelson Memo on the data sharing landscape are unclear, but organizations with stakes in the outcome will need to begin making decisions about how to navigate the new policy now.

Domain repositories could see opportunities for new revenue that can sustain their operations, and should invest in marketing and outreach to communicate their value to potential users. These include being the preferred option of funding agencies and the fact that domain repositories are where the researchers who are most likely to be interested in and reuse data are most likely to congregate. However, the challenges domain repositories will likely face in the new environment are considerable. Their success may depend on their ability to make investments in their infrastructure to ingest and manage larger volumes of data. Well-established not-for-profit and commercial repositories will be better able to raise the capital to adapt to new business conditions than most domain repositories could hope to be. Domain repositories should consider doubling down on their unique ability to serve particular data communities. They might, for example, offer metadata that is more attuned to the needs of a specific community of researchers, develop specialized data management services, or emphasize their status as community gathering places by adding chat features, virtual workshops, or networking opportunities. Some might consider shifting their focus away from archival activity and instead focus on enabling easy discovery of data spread across repositories that is of value to their communities.

The largest not-for-profit generalist repositories have built recognizable brands and a level of financial sustainability that can give depositors confidence that their data will not vanish. Many also have earned the trust of researchers, librarians, and others who will have influence over where data is deposited. Moreover, they have resources that most domain repositories lack to develop new services and features that may provide additional revenue.

3a8082e126