The Challenge of Making Data Publicly Accessible
Jyotsna L Pandey · BioScience · 2016
The push for open science has been gathering momentum since a 2013 White House Office of Science and Technology Policy (OSTP) memorandum directed agencies to make the results of federally funded research, including scientific data, publicly accessible. Federal agencies have since released individual plans to guide the process of making data available to the public. The public-access plan for the National Science Foundation (NSF), like those for several other agencies, includes directives for detailed data-management plans to be required in grant proposals. Many researchers are still uncertain, however, as to how they can fulfill data publishing requirements. According to the NSF plan, which went into effect in January 2016, “data citation and description remain topics of active discussion, and the plan calls for continued consultation with the concerned research communities to formulate guidance on best practices and to incorporate discipline-based innovations as they emerge.” Other agencies, including the National Institutes of Health, Department of Agriculture, Centers for Disease Control and Prevention, National Aeronautics and Space Administration, Food and Drug Administration, and Department of Energy, plan to explore the development of a “data commons.” Such a system would enable data derived from federally funded research to be available for free at the time of publication. But barriers exist beyond the need for a platform to share data. In a letter to the House and Senate Committees on Appropriations in November 2014, John Holdren, director of the OSTP, acknowledged the challenges of changing researcher behavior: “Requiring federally funded scientific researchers to detail how and where they will make their data available in data-management plans is an important first step for increasing access to valuable scientific data, but it is not sufficient to drive the desired sea change in practice across scientific disciplines.” Data availability at the time of publication requires a considerable amount of time and effort from researchers. Some researchers concerned with making their data publicly available argue that this will lead to data scoops by rival researchers who have not invested years in acquiring funding and collecting data. A workshop held in 2014 by the American Institute of Biological Sciences, with support from the NSF, explored how the scientific community will transition to open data. The meeting brought together stakeholders from professional societies, publishers, journal editors, federal government officials, researchers, funders, and representatives of data repositories. From this meeting, it seems that much work remains to be done to develop community-wide practices for making data available and linking them to publications. Workshop participants agreed that journals would play an important role in ensuring that researchers publish their data in a repository at the time of publication. Governments and other research-funding institutions will also have important roles to play in order to ensure the availability of data repositories and long-term funding for maintenance. One successful model is the data repository for the Gulf of Mexico Research Initiative (GoMRI), which is investigating the impacts of the Deepwater Horizon oil spill on the ecosystems of the Gulf of Mexico. All of the data generated during this 10-year, $500-million research initiative must be made available to the public. The GoMRI Information and Data Cooperative (GRIIDC) houses more than 1050 data sets, and the number continues to increase. “The key advantage of having data publicly accessible is to enable and continue discovery in science by having legacy data available for reanalysis at a later time,” said James Gibeaut, director of GRIIDC. “It also makes the scientific process more efficient and economical by allowing researchers to focus resources on other aspects of research rather than on collecting data that is already available.” Gibeaut echoes Holdren's concerns about the need for cultural change: “A major obstacle is developing the culture for data sharing. Getting researchers to realize that this is going to take some of their attention and quality time is a challenge.” A systematic approach to open data is unlikely to be established soon. Pending legislation in Congress ignores the subject. The Fair Access to Science and Technology Research Act would ensure open access to publications from federally funded research but does not include requirements for data. Although directives from policymakers and funders are now in place, much work remains to institutionalize the culture and shared standards required to ensure the success of open data.