Libraries and Open Science: Overlaps and Gaps with the Research Community

This blog post is based on a webinar that explored how libraries and research communities collaborate on open science initiatives. View the recording here.

Academic libraries increasingly play a key role in supporting open science, but what does that support look like in practice? How do libraries effectively partner with researchers, administrators, and other campus units to advance awareness and participation in open scholarship? COS recently brought together librarians and research professionals from diverse institutional contexts to explore these questions.

Defining Open Science: Context Matters

The conversation began with a discussion of how each panelist defines open science, revealing how the term can mean different things depending on one's role and institutional circumstances.

Marla Hertz, Research Data Management Librarian at the University of Alabama at Birmingham, described open science as "an overarching philosophy, and as a movement aimed at increasing transparency throughout the research process," emphasizing both the sharing of research outputs and "creating an infrastructure that fosters those connections between the outputs."

Levi Dolan, Data Librarian for the Health Sciences at Yale University, underscored documentation and accessibility: "Recording the decisions that you make during research, then organizing them and making choices about how you're going to record them… [so] that people, regardless of their context, can broadly interpret what you did.”

“That can involve the choice of tools that you're using, the platforms that you're sharing on, the way that you’re doing publications,” he continued.

Daria Orlowska, Data Librarian at Western Michigan University, brought a unique perspective shaped by her background in human subjects research. Beyond data management and documentation, she emphasized open science as "transforming things that we already have in our collection into data that is now more readily available to people to reuse and to find," highlighting how secondary data can offer researchers a less resource-intensive alternative to collecting new data from scratch.

Tim York, Professor of Human Genetics at Virginia Commonwealth University and Director of VCU’s Data Science Lab, brought up reframing what counts as research products when teaching students: "Maybe think of your publication as an advertisement of your research, but your products are everything"—including data, code, protocols, and manuscript versions. Making these outputs accessible, York said, "is good for science, but it's also good for the researcher, because they make the research more impactful."

Lena Bohman, Senior Data Management and Research Impact Librarian at the Zucker School of Medicine at Hofstra University and Northwell Health, takes a pragmatic approach that varies by audience. "When I present to residents, open science is often the last thing on their mind, even if I'm talking about reproducibility and writing good methods," she explained.

Instead, she noted, framing these concepts around what journals are looking for and how open sciences practices can help researchers get published resonates more effectively. In contrast, bench scientists at her institution tend to be much more familiar with open science terminology, making it a useful framework for those conversations.

Building Strategic Partnerships

As the panelists emphasized, effective open science support requires extensive campus collaboration—and building those partnerships takes sustained effort.

Offices of Research, research compliance, and grants management can be valuable on-campus allies. These offices often run required training, such as sessions on responsible research conduct, that provide pertinent opportunities to introduce open science topics to researchers who may not otherwise seek them out.

York also recommended targeting specific audiences, noting "I've never found that when I shoot a blanket, university-wide email out, even through our Office of the Vice President for Research and Innovation, that many people attend.” But when the message is targeted to a specific department or center, ”the turnout is a lot better."

Similarly, Hertz described the impact of partnering with her institution's Clinical and Translational Sciences Center: "That's a network of over 20 institutions, so when I do a session for them, I'll have 100 to 300 people in that session. I could never get those numbers on my own."

Hertz outlined various approaches to building open science support across campus: synchronous resources on websites and LibGuides; optional workshops and training sessions (ideally, co-promoted with relevant departments), and integration of open science training into required coursework—the latter of which requires ongoing, collaborative relationships with faculty members.

Institutional Culture Shapes Strategy

Institutional context significantly impacts both the culture around open science and the implementation of effective approaches to support it. At VCU, York shared, "Students are very enthusiastic about this movement. They want to be a part of it; they think it's important. They think it's cool."

This enthusiasm helps drive adoption and engage faculty, and as research scales up with bigger data and international collaborations, both students and faculty recognize that open science methods "go hand-in-hand with the increase in scale."

Orlowska faces different challenges at an R2 institution with less federal funding. While there's strong interest in open access publishing and secondary data reuse, broader open science concepts haven't gained as much traction. This, in turn, shapes her approach.

"How do I talk to people coming to my workshops about open science? And how do I make it digestible for them?" Orlowska said, emphasizing the need to meet researchers where their interests already lie, particularly around data discovery and reuse.

At UAB, an R1 with significant NIH funding, the culture is "pretty pro-open access, open science, with some hesitancy on how to do it right." Hertz noted that about half of UAB authors already publish open access, and researchers recognize benefits of data sharing, but worry about doing it responsibly—especially when it comes to participant confidentiality, intellectual property with industry partners, and wanting time for secondary analyses before sharing data.

Reaching Students and Early Career Researchers

The panelists shared several strategies for engaging the next generation of researchers. Bohman found success "framing it around writing and publishing" since trainees are anxious about early publications. They're often surprised to learn that "you can publish a protocol, you can publish a data paper," and seeing open science practices as a way to generate additional publications resonates strongly.

Orlowska recommended focusing workshops on specific tools and skills, then incorporating open science concepts: "Sometimes getting them in the door means… I change my approach a bit."

Hertz underscored the value of reaching out to mentors and training program heads, framing it as: "We can help upskill your students. Otherwise, the mentor is teaching them all these things, like how to find a journal. They can do that, but so can we, and we're saving them that time."

When mentors directly encourage students to attend, participation increases significantly. Hertz and Bohman saw this firsthand when recruiting postdocs for a study—response rates jumped after emailing mentors to ask them to suggest the event to their trainees.

Similarly, York found success with "taking 10-15 minutes" to talk about open science topics at student-organized meetings for graduate students and postdocs, which "usually turns into a longer conversation."

Resources for Getting Started

The panelists highlighted several freely available resources:

• Library Carpentry Lessons: Library Carpentry is a global community that teaches software and data skills to people working in library and information-related roles. They offer courses designed to help librarians support open science, including lessons on Tidy Data, OpenRefine, and the UNIX shell. Lesson materials are available online, under a CC BY license, for self-directed study or adaptation and reuse.
  
  
• Journal Clubs for Open Science: Journal clubs are a low-barrier way to facilitate on-campus and virtual dialogue on open science-related topics. VCU's ReproducibiliTea meetings attract diverse audiences from across campus, with AI topics consistently drawing strong attendance.
  
  
• Additional Resources: Panelists recommended Don't Use This Code (free virtual training on tools like GitHub and Python, and concepts such as FAIR and CARE); the Framework for Open and Reproducible Research Training (FORRT), which offers free resources that are particularly helpful for teaching and mentoring of open and reproducible science; and FASEB's plain-language data management guides.

Moving Forward Together

The discussion revealed both the complexity and the promise of library-research community partnerships in advancing open science. Successful collaborations require understanding the institution's specific landscape and dynamics, building strategic relationships, meeting researchers where they are, and maintaining critical perspectives on what open science means and for whom.

That critical perspective extends to questions of equity and community engagement. Hertz raised aspirational possibilities that extend beyond sharing final research outputs to opening up the research process itself by involving community members and research participants in active decision-making throughout a study, noting that such efforts are particularly meaningful for longitudinal studies and research with human participants—especially "when the makeup of the research team does not reflect the population that's being studied."

Hertz pointed to models like the National Institute for Minority Health and Health Disparities programs—which encourage community members to participate in research decision-making—and the CARE Principles for Indigenous Data Governance, which center ownership with Indigenous populations, as examples of more equitable approaches.

As institutions navigate evolving mandates and expectations around data sharing and public access, strong collaborations between libraries and research communities become increasingly essential for creating research cultures that are truly open, responsible, and equitable.