Sara Laura Wilson, an undergraduate student in the Department of Materials Science and Engineering at the Massachusetts Institute of Technology (MIT), posed this inquiry to 6 materials researchers. Three of the scientists head research labs at MIT and three at Imperial College London.
“Their divergent viewpoints illustrate the challenge of openly sharing experimental materials science data,” Sara says.
Working with Micah Altman, director of the Center for Research on Equitable and Open Scholarship at MIT, and Rafael Jaramillo who leads a research group on electronic materials at MIT, Sara developed a list of survey questions targeting these 3 areas of data management:
- attitudes about sharing data
- data sharing practices
- data stewardship practices
It is indeed daunting to think about how the materials research community will move to data sharing when responses to this small set of questions can vary just among a handful of researchers.
Materials Connect: How did you come to choose principal investigators (PIs) from the Boston area and from London?
SARA: When this project began in February 2018, Professor Jaramillo referred me to PIs at MIT who’ve expressed prior interest in data management and sharing. The project continued over the following year, during which I studied abroad at Imperial College London in the Department of Materials. Professor Jaramillo also had a contact at ICL, through whom I was connected to other faculty at ICL. While all faculty I interviewed performed work related to materials science, their fields of study were very diverse. Through the interviews, I learned of the divergent perspectives on data sharing across materials research.
Materials Connect: What surprised you about their responses?
SARA: All PIs believed that data-sharing should be required, but they had mutually incompatible views of what and how much data should be shared. There was also little uniformity of data management practices beyond the use of Dropbox or a similar database. In general, day-to-day activity was managed by the individual researcher, who did not have set management guidelines, thereby making data searching and interpretation by a third nearly impossible.
Materials Connect: What are a few key outcomes from the survey?
SARA: There are clear needs for improved data management tools, but transitions will remain difficult due to the diversity of hopes, expectations, and current practices in experimental materials science. We need a common set of expectations for data sharing—one that is more nuanced than current requirements but sufficiently general so different researchers could use different software tools to comply. These could be developed by a convening of publishers, materials scientists and their member organizations, research librarians and administrators, and other stakeholders.
Materials Connect: What are your recommendations from here?
SARA: Based on this analysis, data sharing should be comprised of two components. The first is the broad adoption of basic file-level data sharing at the time of manuscript submission. The second component is software for managing data during research. The data management system should require minimal documentation for saved information that can enable data sharing at the time of article submission to become seamless. Despite the groundwork needed for these recommendations, I am hopeful about the future of open materials science. By starting conversations among researchers within and between universities, through both studies like this and other works, the need for data sharing and management systems will be realized. The potential benefits will, hopefully, minimize the burden of undertaking this transition.
Sara and her advisors report on their survey in an article forthcoming in the MIT Undergraduate Research Journal; a preprint, however, is available now: “Methods for open and reproducible materials science.”
-Judy Meiksin
Comments