Data, Decisions, and Doubt: Affording Trust Within Environmental Data Science

Trust is recognised as a valuable element of socio-technical systems, facilitating numerous important benefits. Yet despite extensive study within and across multiple disciplines, achieving well-placed trust remains a perpetual challenge. This research explores the perception of trust within environmental data science: transdisciplinary, collaborative scientific research responding to fundamentally complex, global problems through data science methodologies and techniques. In this context, the presence of trust can reduce uncertainty and friction, consequently facilitating collaboration, generating robust scientific results, and supporting the reuse of data and models. To begin, this research undertakes a theoretical analysis of trust, deconstructing the typical foci of research – models of trust – arguing instead for a different modality: trust affordances. Trust affordances account for trust within system design by attending to multiple stakeholders arriving at trust from varying angles. As a concept, trust affordances are characteristics of the technology by virtue of itself or of features designed into the technology to promote trust by providing access to evidence of (dis)trustworthiness specific to a user, a technology, and their context. This work concentrates on the perceptions of stakeholders across the data science pipeline within two cases. In the first, six interviews and a workshop attended by seventeen participants are used to consider the perceptions of trust and trustworthiness by data curators within a Research Council. In the second case, thirteen interviews are undertaken and a survey of ten respondents are used to generate the views of data producers and consumers related to a Centre of Excellence. The empirical data is analysed twice to respectively produce general themes related to trust and instances of trust affordances. This research contends that whilst there is no universal answer or one-size-fits-all approach to trust, system designers can actively design for trust by embracing nuance and complexity with a creative, thorough, and thoughtful approach.