How can we work together to open science?

Research leaders from government, nonprofits and academia share their challenges and strategies

Open Science panel
The Open Science panel at Elsevier’s Research Funders Summit: Moderator Holly J. Falk-Krzesinski, PhD, Elsevier; Jason Gerson, PhD, Patient-Centered Outcomes Research Institute (PCORI); Carly Robinson, PhD, Office of Scientific and Technical Information (OSTI), US Department of Energy; Sarah Nusser, PhD, Iowa State University; and Mike Huerta, PhD, National Library of Medicine. (Photo by Alison Bert)

If you’re a researcher, you want access to the science you need to do your own investigation – data included and in a form that makes sense. If you’re a research leader, you want to know about the emerging trends in your field and where to focus your resources. If you’re a funder, you want information that will help you compare a research proposal with other research in the field and gauge the impact it will have on society.

No matter what role you play in the research ecosystem, open science is probably important to you. Still, it can take many forms – especially as technology enables new methods of discovering, organizing and sharing data. And if you’re in the business of vetting or analyzing that information, open science can pose a variety of challenges; after all, more is not necessarily easier or better.

Most of us support open science, but how can we make it work for all involved? And how can we do this by working together?

Understandably, open science was a hot topic at Elsevier’s most recent Research Funders Summit. Panelists tackled many of the current and emerging topics related to open science — including open access and preprints, research data sharing, and citizen science — along with incentives to promote open science.

In sharing their own programs and strategies, they mentioned a variety of cross-industry partnerships that help them provide access to research while making sure its properly indexed and linked to relevant data. These included Scholix, which Elsevier helped develop, and CHORUS and Crossref, which Elsevier supports.

CHORUS, which Elsevier supports, enables government agencies, publishers, research officers, librarians and authors to make publicly funded research more accessible.

Making government information easy for the public to find

Dr. Carly Robinson, Senior Product Strategist/Senior Science Advisor for OSTI, shares her organizations methods for making information discoverable and easy to use.At the US Department of Energy’s Office of Scientific and Technical Information (OSTI), public dissemination is a priority, said Dr. Carly Robinson, Senior Product Strategist/Senior Science Advisor. OSTI maintains publicly available collections of scientific and technical information from research, development, demonstration, and commercial applications activities supported by the DOE. Public dissemination is achieved through a set of search tools, including the primary search tool OSTI.GOV , which contains over 3 million records.

“We very much understand that folks might not come directly to our search tools, so we make sure that all of our content is very well indexed, especially in all the (major search engines),” Dr. Robinson said.

Her group uses various methods to make sure information is discoverable and easily usable. These include:

  • Assigning DOIs (digital object identifiers) to DOE-funded technical reports, workshop reports, datasets and software. They collaborate with Crossref and DataCite for this purpose.
  • Interlinking data with scholarly literature. To do this, they take part in Scholix: Framework for Scholarly Link Exchange, which Elsevier has also been very active in.

The end goal, she said, is to interlink all related research results: for example, the publications to the software to the data.

An example of how OSTI interlinks research results.

Helping people make informed healthcare decisions (while protecting patient data)

Dr. Jason Gerson, Senior Program Officer for Clinical Effectiveness and Decision Science at PCORI and Co-chair of the Open Science Task Force of the Health Research Alliance, talks about the importance of putting data in context.The Patient-Centered Outcomes Research Institute (PCORI) deals with its own set of challenges as a funder of clinical comparative effectiveness research (CER). Clinical CER compares healthcare treatments to determine which work best for which patients. PCORI funds clinical research studies that help people make informed healthcare decisions by producing evidence guided by patients, caregivers and the broader healthcare community.

Because PCORI-funded studies often generate individual patient data, their data sharing policy (approved in September 2018) requires that all such data be de-identified in accordance with HIPAA regulations. Additionally, participants must give informed consent for their data to be used for research purposes. The policy also requires that all requests for access to the data be vetted by an independent review panel for scientific merit.

Importantly, PCORI’s policy focuses not just on the dataset but the data documentation that is critical to making sense of the dataset.

As Dr. Jason Gerson, Senior Program Officer for Clinical Effectiveness and Decision Science at PCORI, explained:

It can’t just be the dataset – it has to be a full roadmap, which includes the study protocol, the metadata, the data dictionary, the statistical analysis plan, and the analytic code.

His organization focused on creating a data-sharing policy that was clear to all involved, including PCORI colleagues, awardees and people outside the organization requesting data. Features of the policy include specifying the data and documentation to be shared, outlining the expectations for data management and data sharing for awardees, and providing funding to support awardee’s efforts in preparing data.

Features of PCORI’s data-sharing policy.

“Living at the intersection of data science and open science”

Dr. Mike Huerta, Associate Director for Program Development and Coordinator of Data Science and Open Science Initiatives at the National Library of Medicine, talks about resources, tools and services provided by the NLM. (Photo by Alison Bert)

Over the years, the National Institutes of Health (NIH) has led many large-scale data-centric and open research initiatives, including the Human Genome Project and the Cancer Moonshot initiative.

A sample of resources led by the NIH.

A key part of the NIH is the National Library of Medicine (NLM). As the world’s largest biomedical library, it creates and hosts major resources, tools and services for literature, data, standards and more, including PubMed, PubMed Central and GenBase.” These resources are all free for the public to access online.

“I think of NLM as living at the intersection of data science and open science,” said Dr. Mike Huerta, Associate Director for Program Development and Coordinator of Data Science and Open Science Initiatives at the NLM. “I see this as the most promising space right now for the advancement of biomedical research.”

Dr. Huerta said various factors were converging to make biomedical science more data-centric and open, including societal expectations, policy, technological advancement and scientific opportunities.

Various factors are converging to make biomedical science more data-centric and open.

How can funders help researchers and universities share data in a meaningful way?

Prof. Sarah Nusser, PhD, of Iowa State University talks about accelerated public access to research data.

Universities are working to implement data-sharing services that enable researchers to make their data accessible to other researchers and the public while complying with agency mandates. However, they face serious challenges, according to Dr. Sarah Nusser, Vice President for Research at Iowa State University and Co-chair of the AAU-APLU Public Access Working Group. For example:

  • Research data do not arise from a common process designed to generate a discrete final product.
  • Many disciplines do not have an established sharing practice.
  • Data sharing and open research methods are entirely foreign to many researchers.
  • Underlying systems, policies, research practice guidance, and compliance workflow are underdeveloped.
  • Funder requirements are diverse and often vague.
  • Resources are inadequate to implement and offer new services to support data sharing.

To address these challenges, the AAU-APLU Panel on Open Science created a roadmap with suggestions for funders and universities:

The AAU-APLU Roadmap includes these principles for open science.

Suggestions for funders

  • Harmonize policies, documentation and requirements across funders.
  • Develop a minimum standard for data access requirements (e.g. data for publications), with exceptions for disciplines.
  • Specify compliance approach for monitoring, evaluation and enforcement (e.g., how it will work).
  • Prioritize quality and value of data and its documentation.
  • Balance costs and benefits in setting retention requirements.
  • Work across funders, universities and professional organizations to develop standards, common methods and interoperability.

“If you did one thing on this list,” Dr. Nusser said, “find a way we can all harmonize what we’re doing to help researchers be successful. Even if there are a few core elements that are common across sponsors and institutions, that will do a lot to reduce the burden faced by a research office in supporting researchers.”

Suggestions for universities

  • Focus on a minimum standard for data access requirements (e.g., data for publications), with exceptions for disciplines.
  • Support researcher practice through policies, guidance and training as well as services and systems to support workflows for sharing data.
  • Reward researchers for providing public access to high quality and impactful research data.
  • Work across funders, universities and professional organizations to develop standards, common methods and interoperability.

“We need to learn how to reward researchers better,” Dr. Nusser said. “We need to reward them for lots of things, but high-quality, impactful research data is one of them.”

Quick question for you

Which terms do you most associate with Elsevier? (check all that apply)

Data and analytics
Research platforms
Technology
Decision support tools
Publishing
Books and journals
Scientific articles
Healthcare content

Tags


Contributors


Comments


comments powered by Disqus