In our most recent class discussion (March 17, 2016) we had the good fortune to host two guests.
Dr. Terry McGlynn, Professor of Biology at Cal State Dominguez Hills (and contributor to the blog http://www.smallpondscience.com)
Dr. Emilio Bruna, Distinguished Teaching Scholar and Professor from the Department of Wildlife Ecology & Conservation & Center for Latin American Studies (contributor to http://brunalab.org/blog; Editor and Chief of Biotropica)
Up until this discussion we had primarily covered the nuts, bolts, best practices, authorship assignments, and advantages of data management and sharing. But yesterday we took a step back to consider some of the reasons (beyond ego or stubbornness) that much of the scientific community still holds reservations about releasing hard-won scientific data. Both Emilio and Terry are involved in the open science community, and I think I am safe describing them as very open-minded skeptics of the increasing push to require scientists to make data publicly available. Their unique perspectives and experiences as researchers brought some very interesting issues to the fore. The conversation covered a lot of ground but I will highlight a couple points that struck me the most.
Checking our science privilege
Both of our guests made very interesting observations from their unique perspectives as researchers. Terry, who works at a teaching institution, conducts research in the face of some major challenges compared with his peers, many of whom work at much larger R1 universities. For one, a heavier teaching load and reliance on undergraduates to conduct research can make the publication process slower compared to a researcher with a group of graduate students and postdocs and low teaching responsibilities. For this reason the one-year embargo period, which is often advocated to give researchers a chance to have exclusive access to their data for a limited time, makes less sense for Terry than it would for individuals at research institutions. Moreover, the greater relative time cost to Terry for data collection (i.e. on top of his teaching and service responsibilities) makes his data more valuable to him. Furthermore, collaborations are harder to come by at smaller colleges. Making his data freely available might limit his opportunities to collaborate on projects with other researchers who would not have to interact with him to access his data. Although one could debate these costs of data sharing in general, it seems clear that researchers at smaller universities would be hit harder by being forced to relinquish data than individuals at larger universities.
Emilio’s perspective as a Professor of Latin American studies and collaborator with many researchers in developing Latin American nations highlighted some of the extra costs that researchers from these regions would experience from relinquishing control over their data. Many of these disadvantages would be similar to those experienced by Terry at a teaching institution, which increases the argument that the costs associated with open science will not necessarily be equitably distributed. He posed an interesting question about authorship vs. collaboration. Is offering authorship to a data contributor to a metadata study, but not offering an opportunity for collaboration, really all that fair to the contributor? I wonder, what if the individual in a developing country did not have the same data access opportunities to publish a metadata study, but still had to fork over data? This strikes me as almost abusive. We concluded by agreeing that the system of credit is certainly in flux.
Does long-term data enjoy a special status?
There are some important questions here. Does long-term data represent a special situation? Many long-term studies were initiated long before most researchers were imagining future data sharing capabilities or the open science movement. Researchers who work very hard to contribute to these datasets sometimes view their effort as an investment in the future where they expect to reap the benefits of the data toward their own careers. If the data are continually released, this could potentially devalue the contribution. For comparison, shorter-term study datasets are typically not released piecemeal as the data are created. This raises the question: will researchers be dis-incentivized to initiate and participate in long-term data collection without the promise of the long-term benefits? My inner Homo economicus says “yes,” but my gut says that “if the research question is cool enough and funds are available, then no.”
Thanks again to Terry and Emilio! Go check out their blogs and lab websites! 🙂