He won a travel grant from JoVE to attend the Charleston Conference for his leadership in promoting research reproducibility on his campus and to the librarian community. Listen to Franklin present his work on the panel ‘Librarians Leading the Way to Improved Research Reproducibility’ on Nov. 8th in Charleston.
Ed. Note: This interview has been lightly edited for clarity.
JoVE: Why is Reproducibility a key issue today in academic research?
Franklin Sayre: Reproducibility is a core part of the scientific method, and the current reproducibility ‘crisis’ has consequences for the extent to which our students, faculty, and the public at large can trust experimental results. Our faculty builds on these results with their own research, and our students and the public use these results for everything from public policy to patient care.
J: Do you think librarians ought to get involved in this process?
FS: As information experts, librarians have a responsibility to help users understand systematic biases and problems with the scientific literature. Academic librarians are also uniquely positioned to help with reproducibility. We have broad representation across institutions, including liaisons embedded in each discipline. Many of the recommendations for improving reproducibility are core areas of academic librarianship, including data management and sharing, scholarly communication, and methodological support for systematic reviews and data-intensive research.
J: Are librarianship and reproducibility aligned today?
FS: Issues with research reproducibility are starting to gain visibility within librarianship, but we have a long way to go before it becomes a routine part of how we frame our work. We need to convince librarians that reproducibility is directly related to our services and expertise, and we need to convince our users that librarians have much to contribute to solving systemic problems with reproducibility. Finally, we need to offer pragmatic ways that library services and librarian expertise can directly impact reproducibility that don’t require every librarian to learn about the topic from scratch.
J: How did you personally get involved?
FS: As a medical librarian I’ve taught evidence-based medicine and critical appraisal to clinicians and have known about issues with the research transparency, rigor, and bias for years. When reproducibility started gaining widespread attention after the Reproducibility Project I thought the time was ripe to start thinking about how these issues impact both how we support researchers and how we teach students who use research. My current focus is on thinking broadly about how our services and expertise can contribute to reproducibility. I’m also working on a proposal for supporting automation of data processing and analysis in order to improve reproducibility.
J: Can you provide examples of how you put your services to use?
FS: In my areas of responsibility, I have developed many innovative ways to approach reproducibility. I worked with a colleague to develop an educational session for laboratory meetings that taught active lab data management using the lab’s actual data with the focus on immediately improving description and preservation. I also initiated a group that looked into Electronic Lab Notebooks and their use on campus and have worked with five large biomedical labs on evaluating and selecting ELNs. This work led to a collaboration with a research quality assurance office on campus to submit a successful supplementary NIH training grant for training graduate medical scientists about quality assurance and reproducibility.
J: What advice do you have some other librarians starting with reproducibility?
FS: Think carefully about the boundaries between our expertise and services, and the discipline itself. This work is all boundary work – we are pushing the boundaries of traditional library expertise – so it’s important to be thoughtful about that and how we fit into the picture. Finally, partner with others who have expertise that we don’t, such as statistical support units, where possible.
J: Are there any trending tools or content you would recommend?
FS: A lot of the most exciting developments in reproducibility are about workflows and automation. Tools like Open Science Framework, electronic lab notebooks, R, and other open computational tools provide scaffolds for busy researchers who want to address reproducibility within their research. Preprint servers, repositories and pre-registrations are also essential for broadening access to information about research in order to improve reproducibility and rigor in research.
Want to receive some useful reads and resources about reproducibility?