Form Submission: Session, Poster, Demo Submission

Demo/Poster Number


Contact Information
First Name: 
Last Name: 
Session Information
Presentation Title: 
How scientists assess trustworthiness of data produced by others: Do practices shape infrastructure, or vice versa?
Scientists in many domains increasingly need to use data produced by others in their own work. They may be using data produced by collaborators working in different geographical locations, they may be reusing data produced by other scientists for other purposes, or they may using data produced by large projects such as astronomy sky surveys or the Human Genome Project. Frequently, the potential data user data will not have witnessed the production of these data; instead, they must make a judgment that the data is trustworthy. Studies of how scientists trust in data produced by others reveal that scientists employ one (or more) of a range of strategies to assess data trustworthiness, and that these strategies can vary by scientific domain. Sometimes, potential data users scrutinize the methods used to produce data; sometimes, they assess the competence and integrity of the data producer(s); and sometimes they make judgments about the large-scale system that produces the data. The importance of building infrastructure for scientific data that incorporates the information necessary for scientists to assess data trustworthiness is well recognized. Typically, the relationship between trust practices and infrastructure is conceptualized as identifying the strategies that scientists in a particular domain use to assess trustworthiness, and then building infrastructure that provides information relevant to this strategiy. Drawing on a long-term case study of scientists who study microbial life in the seafloor, I invert this relationship. The strategies of scientists in this domain are shaped by the infrastructural resources already available: if resources are not available to assess trustworthiness according to one strategy, then scientists apply a different strategy for which resources are available. Developers of data infrastructure should consider not only how to support existing strategies, but also that the infrastructure they develop may reconfigure scientists’ strategies for assessing data trustworthiness going forward.
Presenter Information
Same As Contact: 
University of Illinois at Urbana-Champaign