Capitol Corner — July

Capitol Corner — July

Published monthly as part of FLC’s DC Perspective content, Capitol Corner focuses on one notable news item pertaining to the T2 community. The focus stems from agency publications, news sites, and DC-central organizations, with original sources, contacts, and links provided. For more information and Corner-related inquiries, please contact dcnews@federallabs.org.


After a series of symposiums and meetings, the National Academies of Sciences, Engineering, and Medicine (NASEM) published a consensus study report, “Open Science by Design – Realizing a Vision for 21st Century Research,” earlier this month. NASEM’s report was assembled from an ad hoc committee under the Academies’ Board on Research Data and Information (BRDI) in order to pave the way for a research and development (R&D) future built on open science. Open science champions a research landscape where all publicly funded research items (e.g., scholarly articles, code, algorithms, etc.) are made findable, accessible, interoperable, and reusable (FAIR) to the public, provided the release of these items poses no threat to privacy, trade secrets, or national security.

The report is designed around several recommendations from the Committee on Toward an Open Science Enterprise for how the research community can adopt this new philosophy, which creates an “open science by design” framework that is threaded through each stage of the research process. Let’s take a closer look into the report’s findings and potential ramifications.

Open Science Today

Open science and access has largely been an international affair. In 2002, the Budapest Open Access Initiative (BOAI) called for free and open access to scientific literature, and the European Commission has built policy language around the creation of a European Open Science Cloud this year. In the U.S., a 2013 memo issued by the Office of Science and Technology Policy (OSTP) called for R&D-heavy agencies to develop plans to encourage open access to research data and the publication of data management plans. Current American implementation of the OSTP memo’s request includes the National Institute of Science and Technology’s (NIST) participation in the National Data Service (NDS) project. The NDS is inching toward the creation of data portals that cut across scientific disciplines and align to FAIR principles—helping researchers find, use, share, publish, and link data. Other current open science initiatives include open access journals and repositories, policies at research universities, and those implemented by private institutions, publishers, and professional societies.

Introducing the Open Science Framework

The “open science by design” framework is designed with the overarching principle that research conducted and published openly and transparently leads to better science. Although NASEM claims that this framework is “by necessity general and idealized,” the report claims it benefits researchers who share data and champion research collaboration. The framework is broken into six steps, outlined below.

  1. Provocation. Because open science promotes open research, researchers have immediate access to FAIR publications and information. Researchers explore these resources to identify new concepts to explore and identify where new contributions can be made. Researchers can also network with colleagues to build new hypotheses and experiments.
  2. Ideation. After researchers’ contributions have been identified, research plans are developed and pilot studies conducted using preliminary data from FAIR repositories.
  3. Knowledge generation. Researchers collect fresh data using tools compatible with open sharing, including automated workflow tools, to ensure the accessibility of research output.
  4. Validation. Data is collected, analyzed and interpreted. At this stage, replication studies refine the research approach to ensure reuse and replication. When revisions are complete, data is researched in standard formats according to disciplinary standards.
  5. Dissemination. Peer reviewers offer comments to improve pre-published research prior to open publication of data, code, articles, and other artifacts. All research is published under a public copyright license (e.g., GNU General Public License, Creative Commons License).
  6. Preservation. Final research artifacts are deposited in a FAIR archive so researchers can repeat this process to build on the published work, starting at Provocation.

Benefits of Open Science

The report highlights several motivations to transition open science into mainstream adoption, based upon how open science has succeeded today. These benefits include those below; for the full list, review the report highlights here.

  • Rigor and reliability. With code and data being made publicly available for previously reported work, studies are easier to replicate, which strengthens scientific reliability.
  • Ability to address new questions. Data under open science is made accessible from a wide spectrum of scientific fields, which allows researchers to collaborate across disciplines and areas of inquiry.
  • Faster and more inclusive dissemination of knowledge. Because open science accelerates the accessibility and subsequent dissemination of research, more researchers can participate and expand upon published results, thereby encouraging worldwide research collaboration.
  • Improved performance of research tasks. Tools like electronic lab notebooks automate data collection and the accurate recording of experimentation, which allows for streamlined research processes.
  • Open publication for public benefit. Open science benefits taxpayers as this accessible funding can be used more rapidly to improve health, protect environmental quality, and deliver new products and services.

Barriers to Open Science

Despite the benefits, open science as envisioned has experienced some roadblocks. These barriers include those below; for the full list, review the report highlights here.

  • Costs and infrastructure. Widespread open publication and open data dissemination require new technological and institutional infrastructure across disciplines, which needs to be developed.
  • Lack of supportive culture, incentives, and training. Current incentive and reward systems do not champion code and data sharing, as tenure is awarded to those with ownership over certain research.
  • Disciplinary differences. In many research disciplines, there is no universal standard for the treatment of data and code, and dataset size might affect the robustness, speed, and effectiveness of data sharing and the representation of it across disciplines.

Recommendations

After the conclusion of this study, NASEM’s Committee suggested the following recommendation to break down the barriers listed above. (Note: Those not listed above but included in the report highlights are deliberated on in greater detail in the final report.)

Building a supportive research culture starts with retooling reward systems and creating open science training opportunities. Universities and other research facilities should reward those engaged in open science activities. Research funders should also reward the use of data available in public repositories gleaned in the Provocation stage and archived in the Preservation stage. In order to understand the importance of open science, universities and research funders should support the creation of training programs in open science best practices and the principles of the “open science by design” framework.

For more information on this report, visit NASEM’s site, which includes a link to the report release webcast, an accompanying press release, and other supplementary materials.

Category: 
DC on T2