Open Up Engineering Education Research
It’s time for a dialogue on sharing research data.
Opinion by Aditya Johri
In March, 2015, the National Science Foundation released its Public Access Plan (Today’s Data, Tomorrow’s Discoveries), which lays out an impressive agenda to improve transparency of public funding by increasing access to research outcomes and data. The Public Access Plan is a necessary and ambitious document and follows in the steps of similar initiatives by the National Institutes of Health and other agencies. It is important for engineering educators to pay attention to this policy not only because the community receives significant support from NSF but also because the release of this document provides an ideal opportunity for the community to engage in an honest dialogue around a persistent problem in the community – the lack of sharing of research data.
As part of a recent research study on data sharing, funded by NSF, I have had numerous conversations with engineering educators who have all expressed their frustration with the lack of data sharing within the community. They contend that it’s the ethical and moral duty of a researcher to be as open as possible with research funded by taxpayers. They also argue that the lack of shared research data can lead to duplication of research efforts and resources (financial and physical), an inability to replicate or provide alternative explanations for findings (both hallmarks of peer-reviewed research), and exclusivity, as resource-poor researchers and institutions cannot participate in the research enterprise. Almost everyone I spoke to expressed an interest in crafting a common charter for more open and efficient research practices within the community, including shared norms around access to data.
Even though interest in data sharing is high, researchers are keenly aware of potential barriers. They understand now that there is a need to better understand institutional review board-related issues, especially concerns with protecting the privacy of participants. Data are often highly dependent on the context for interpretation, and lack of description or meaningful metadata can make sharing unusable. Therefore, solutions such as sharing of de-identified data are easier to achieve with quantitative data but not necessarily with qualitative data collected for interpretive research.
Community members also cited the absence of incentives for data sharing. For instance, research funding is directed towards new data collection rather than secondary data analysis. Within higher educational institutions, the tenure and promotion process values external funding and publications, but efforts to share data or create similar infrastructures go unrewarded.
Some researchers are reluctant to share data because they plan to publish it and fear somebody else might get the benefit.
Respondents identified several ways for data sharing to be supported. They expressed a desire for an easy-to-use online system that would require minimal effort and allow for uploading, downloading, and managing permissions. Best practices for data management throughout the research process could be integrated with data sharing. Such integrative workflows are common across other fields, and as a maturing research field, engineering education is well positioned to develop them. Educators suggested that incentives be designed to encourage both data sharing generally and secondary data analysis. Protocols are needed to guide who can publish first using shared data as well as authorship credit.
Sharing information is essential if institutions of higher education and their faculty members are to remain at the forefront of knowledge creation. If we want to make the engineering education field more inclusive, we need to reduce barriers to entry. A community effort to share research data is a critical step in that direction. Finally, technological advances will help, but they have their limits; research is a socio-technical enterprise, and norms have to come from within the community. In this effort the community can find models in such successful projects as the Inter-university Consortium for Political and Social Research, Qualitative Data Repository, and Databrary, an open library for sharing video data. It is time to have an honest dialogue.
Aditya Johri is an associate professor of information sciences and technology and director of the Engineering Education and Cyberlearning Laboratory at George Mason University. He co-edited, with Barbara M. Olds, the Cambridge Handbook of Engineering Education Research. This work is partly based upon research supported by U.S. National Science Foundation (NSF) Award # EEC-1408674. Opinions expressed do not necessarily reflect the views of NSF.