Philosophers of science have long held that scientific data have a special status in science. They represent facts in the world (Hempel, 1952), serve as evidence for claims about phenomena (Bogen & Woodward, 1988), and function as arbiters for theory testing (Schlick, 1935). Contemporary scholars highlight that the ways that data are produced, used, and handled shape how they serve as evidence (Borgman, 2015; Kitchin, 2014; Leonelli & Tempini, 2020). Separately, there is now a growing consensus …
Read morePhilosophers of science have long held that scientific data have a special status in science. They represent facts in the world (Hempel, 1952), serve as evidence for claims about phenomena (Bogen & Woodward, 1988), and function as arbiters for theory testing (Schlick, 1935). Contemporary scholars highlight that the ways that data are produced, used, and handled shape how they serve as evidence (Borgman, 2015; Kitchin, 2014; Leonelli & Tempini, 2020). Separately, there is now a growing consensus among philosophers of science working on values and science that values influence many scientific decisions. In addition, they suggest various proposals for distinguishing between legitimate and illegitimate influences of values in science (Brown, 2020; Douglas, 2009; Elliott, 2022, 2017; Intemann, 2015; Steel, 2015). To combine these two pieces of literature, this thesis provides an examination of how values guide and impact data practices with an emphasis on data sharing and reuse. In addition, I explore how these values can be managed to develop socially responsible and epistemically beneficial practices. Despite the numerous examples of value-laden practices, data practices have not received much attention in the literature. I examine how values can influence many data practices including data collection, sampling, analysis, sharing, and reuse. Then, I discuss how philosophical proposals can help identify, manage, and justify values in data practices. In doing so, I scrutinize individualistic proposals that portray individual scientists as primary decision-makers in value management (Brown, 2020; Steel, 2015). Then, I will discuss how value management can be enhanced by restructuring research communities (Longino, 1990), stakeholder engagement (Intemann, 2015), and democratic decision-making (Kitcher, 2011; Lusk, 2021; Schroeder, 2021). Lastly, I will offer an institutional approach by highlighting the roles of institutions in value management such as determining rules, norms, and principles for research, identifying and promoting values, as well as facilitating community engagement. I defend a pluralistic account of value management by which values can be managed in at least three distinct but complementary ways: identifying and selecting values, consensus, and democratic selection of values. To reflect this pluralism, I explain how different institutional strategies that can be used to manage values in different ways.
