Lutishoor (Luti) Salisbury is a science librarian and head of the Chemistry and Biochemistry Library at the University of Arkansas.
She won a travel grant from JoVE to attend the Charleston Conference. She will share her deep understanding of the issue of research reproducibility and the role librarians have to play on the panel ‘Librarians Leading the Way to Improved Research Reproducibility’ on Nov. 8th in Charleston.
JoVE: Why is reproducibility a key topic in science?
Luti Salisbury: Reproducibility of science is important today because of the serious and widespread shortcomings and biases regarding the replication of studies have been reported and discussed in published papers.
J: Do you believe reproducibility is librarian’s business?
LS: In a world of proliferation of fake news, diverse information sources, the pressure to publish or perish and the desire to gain tenure at higher education institutions, it has become increasingly important for librarians to take an active role in engaging their faculty, researchers, and students on the best practices of research transparency and reproducibility.
J: How can librarians play a role in building awareness on the topic?
LS: I believe that as a science librarian I have a moral responsibility and a huge role to play in disseminating the understanding of the research life cycle, how information is fed into each stage, and how the results of experiments or research should be made robust, unbiased, findable, reproducible and usable.
Librarians reach out to a large number of researchers on campus in instruction sessions and at the reference desk, so they are uniquely positioned to bring these studies and its surrounding issues to the attention of the faculty, graduate students, and other researchers on their campuses. This will bring attention to the issues in classroom sessions, labs and in group meetings. We are also uniquely placed to advocate for the public availability of published research data.
J: What areas of library science overlap with research reproducibility?
LS: The librarian skillsets that are most valuable to improve reproducibility are: knowledge of the scholarly communication ecosystem, an understanding of how scientific research is conducted, understanding of the statistical methods that are used to draw conclusions; being able to teach how to critically evaluate information, how to acquire skills in data management, understanding storage and retrieval to aid in replication; finding all the corrections, refuted articles, and retractions related toa particular paper.
J: How do you apply these skills in your day-to-day work?
LS: In my instruction sessions to graduate and undergraduate students and other researchers, I always make an attempt to introduce the scientific communication system, talk about the scientific process and show it should be objective and rigorous, even though this is not always the reality. I teach students how to find reputable material, how to evaluate sources for biases and fake material, and how to follow up to find corrections or letters to the editors that may refute or show a different dimension to the claims made in a paper. I help faculty in their data management efforts. I also teach patrons how to find the raw data relating to a paper which may aid in replication.
J: Can new resources and tools help scientists make their research more reproducible?
LS: The reproducibility crisis applies to two areas relating to the scientific paper. The first is in reproducing the method in the paper as outlined by the author. In describing the method, if the author misses a tiny step in his narrative it may make the experiment difficult to repeat. As a result, much time and effort can be wasted trying to repeat the experiment without results. By taking advantage of the new technologies, such as JoVE and others, to film the method live in the lab, there is less chance of missing a step in the process. If these are attached as supplementary information to the particular paper it will be easy to find and use to repeat the experiment.
The second is in using the data upon which the results of the paper are based to verify/repeat the experiment. Technology can be very useful to store, retrieve and manipulate large quantity of data to aid in reproducibility.
Want to receive some useful reads and resources about reproducibility?
Ed. Note: This interview has been lightly edited for clarity.