On Friday, March 21, members of the science and policy communities met to discuss the current challenges and opportunities in neuroscience data sharing as well as possible ways to advance data sharing going forward. Panelists included representatives from the neuroscience, industry, statistics, database, funding, and scientific journal communities. The biggest obstacle facing neuroscience data sharing appears to be the need for cultural change (i.e. creating an environment in which data sharing is part of the work flow for scientists instead of an afterthought and a burden). One solution that was discussed is the inclusion of data sharing practices in the evaluation of promotion and tenure decisions. Another is the development and widespread use of technologies that make data sharing an easy part of the scientific process. A second obstacle is the high cost for not only sharing data, but also maintaining and curating that data. It is currently unclear whose responsibility this should be. Before sharing can become widespread, the community must determine exactly what data should be shared. There are differing opinions on this; should investigators share all raw data, only processed data, or only data pertaining to the experimental questions of the study? Improvements in hard drives continually make it easier to store and share large quantities of data, but as more and more data is collected with new tools and techniques, this will become a bigger issue and one that must be addressed.
As science becomes an even more international venture, the melding of international data sharing policies will also become increasingly important to allow for data from around the world to be accessible to all. Discussion at this one day symposium revealed various issues facing sharing of neuroscience data which, once addressed further, will contribute to the creation of new policies in the neuroscience community in order to lower the costs and obstacles to data sharing and foster an environment where data sharing is encouraged and valued.