Paper short abstract:

Epistemic and non-epistemic values driving data sharing (data access) governance and practice illustrate the complex orientations of data generators, researchers and others to open science; in particular, protecting the participant, protecting the study, and protecting the researcher.

Paper long abstract:

Open science, in the context of bioscience studies involving human participants, faces inherent contradictions. Sharing data from research studies is widely understood as essential to meet the statistical power necessary to produce valid findings about biological processes with small effect sizes, eg in genetics, epigenetics and 'omics' research. But bioscience data can never be fully open if the expectations of privacy and confidentiality of research participants are to be maintained.

Drawing upon the ethnography of a large European consortium aiming to develop new technologies for data sharing and harmonisation - BioSHaRE-EU - we present an account of data sharing in practice. Particularly, we studied the pilot of an integrated mechanism - comprising DataSHaPER, DataSHIELD, Mica and Opal software - offering secure privacy-protecting analysis in which individual data remains within the firewall of the data-generating study. Based on interviews with study PIs, developers of the technology and researchers wanting to access study data we present analysis of epistemic and non-epistemic values driving data sharing (data access) governance and practice. Three values illustrate the complex orientations to open science: protecting the participant, protecting the study, and protecting the researcher. An "old fashioned" science led by singular individuals was juxtaposed with a "new 'open' science" characterised by collaboration among and beyond scientific stakeholders. Yet those values were in flux, with some individuals fully reversing their position towards openness during the five years of the consortium. Governance practices were slower to change and current forms of governance are struggling to address the incongruities of open (bio)science.