Data Sovereignty and a University Partnership
A university research team proposes a five-year study of chronic illness in a First Nation of 1,100 people. The team offers $40,000 per year, two community research assistants, and co-authorship for a designated community representative. The draft data-sharing agreement says de-identified data will be stored on university servers and made available to 'qualified researchers' on request. The health director notes that a previous study collected blood samples in the 1990s; those samples were later used for research the community never approved, and results were published without community review. Community members remain angry about it. The band's own health survey, run by two community members, already has 300 responses.
1. Which provisions in the draft agreement should the community seek to change, and why does each matter?
Read the worked answer
Three provisions are problematic. Storage and access: 'qualified researchers on request' hands control of community data to people the community never vets. Given the 1990s history, this is the provision most likely to trigger community opposition. The community should require that data remain under community control or joint control, with a governance body approving any secondary use.
De-identification as the model: de-identification protects individuals but not the collective. In a community of 1,100, small-cell data can effectively identify families or the Nation itself. The agreement should specify how results are aggregated and reviewed before release.
Review of publications: co-authorship for one representative is not the same as community review. The community should require a review period before submission and a right to contextualize findings. The earlier harm was not only sample misuse but publication without community voice; the agreement must address both.
2. How should the health director use the existing community survey in negotiations?
Read the worked answer
The 300-response survey is leverage. It shows the community already generates its own data and is not dependent on the university for basic information. The director can propose that the university study build on and cite the community survey rather than duplicate it, which reduces burden on participants and positions the community as a data owner, not just a data source.
The director should also ask who holds the survey data and under what terms, and propose that the partnership agreement cover both datasets under one governance framework. This turns a defensive negotiation into an offer: the university gets a richer baseline and community trust; the community gets control. If the university refuses governance terms, the survey gives the director a credible alternative - continue community-led research and decline the partnership.