And I am not talking about the stars in Hollywood…
I was in a conversation with several colleagues the other day, discussing the barriers that get in the way of people sharing data with others. Everyone in the room held the same hypothesis: more data sharing would lead to more analysis, more insights, more problems solved, and ultimately greater impact in the fields in which we work. The purpose of our conversation was to determine whether we had a common definition of “data sharing” and to scope the problem, in order to explore what action we might take. One of my colleagues spoke about his former field, astronomy. Apparently, there are less issues with data sharing in astronomy, due to a few important factors:
- There are very few funders of the work – mostly the federal government and a small number of large consortia – so they can exert a lot of influence regarding policies and expectiations
- There is one primary platform for data collection – the Hubble Space telescope
- Anyone can access the data collection method, through an application process, and all data collected through the method are made available to anyone on the internet after a set period of time, making it available for additional analysis
When we thought about how the experience of astronomy applied to other fields of interest, we landed on three potential levers we could act on to create change:
- Tools or solutions that make sharing data as easy as possible
- Structural supports, like policies and procedures
- Incentives, such as rewards for sharing data
Our group explored this through reflection on another example where at least two of the three interventions were put in place in order to increase sharing of data in a particular field: a common platform to store data, policies such as a common data standard, and an API layer implemented to increase access to the data in the platform. The program saw initial success, and things were looking positive. And then it ran in to issues, primarily with the lever of structural supports: regulatory blocks in the public sector, and privacy concerns regarding the use of the data itself. The privacy concerns alone were enough to undermine political support for the program. Clearly, privacy is an important issue that needs to be addressed – and any field in which that particular concern does not exist means that there is one significant barrier removed for greater access to data. Since stars don’t care about their personally identifiable information, astronomy does not have to address that along with other issues.
Of course, these are only two examples. And it seems like a reasonable hypothesis that road blocks in any of the three areas listed above could easily slow forward movement toward greater sharing of data, and that multiple issues could stop progress all together. I think it can be difficult to determine how best to intervene to create change, because the issues in each of the three areas are often inter-related in a complex system. Using a system dynamics frame to understand the interrelationships may help, which we are currently doing. I hope to share insights we glean from this process on this blog in the future.