1 Comments
NIH supports wants your input on strategies for data management, sharing, and citation – the what, when, and how of data management and data sharing, as well as the standards for citing shared data and software. As described in a recent blog post by Dr. Carrie Wolinetz, NIH Associate Director for Science Policy, to move forward with our ongoing commitment to data sharing in the scientific enterprise, NIH wants to prioritize the many facets of data management and sharing, and consider these priorities in the context of existing data sharing policies, such as the 2003 NIH Data Sharing Policy.
Check out the links below to learn more about the background and details of this RFI, including how to submit your responses. While general comments and discussion are welcome in the comment field below, be sure to follow the RFI instructions for submitting your comments by the December 29 extended January 19 deadline for NIH’s formal consideration of your feedback.
[Note: The response date has been extended: Electronic responses will be accepted through January 19, 2017. –NIH Staff, 12.15.16]
Learn more:
“Under the Poliscope” blog: The What and How of Data Sharing
NIH Office of Science Policy website and RFI submission form
I have no problem with data sharing per se. However, there is considerable variability across institutes in terms implementation. There are also onerous bureaucratic requirements. These issues can impose significant time burden on scientists who try to comply with the requirements, thus detracting from scientific progress, especially where the data thus collected are of dubious quality and utility. I offer up as an example sharing of genomic data in the field of autism genetics. The genomic data is stored not by dbGaP, whos staff knows how to collect and store genetic data, but by NDAR at CHDD, who re-invented the wheel. The result has been storage formats that have been known to be problematic for >80 years in the field of human genetics. The formats are also extremely difficult to create, given the data that need to be aggregated in order to do the data submission. Part of the problem seems to be the inability of units at NIH to communicate together (e.g., exchange appropriate subject IDs to store different aspects of data). Rather than dumping this problem on the extramural research community by storing data in an increasing number of separate NIH databases, NIH should get its act together and figure out how to make it possible to store data so that types of data collected by funding from more than one institute have a single repository for those data (e.g., genomic data goes to dbGaP, and *only* to dbGaP), with a single interface that investigators need to learn to use.