Educating graduate students aims at making them competent members in a disciplinary community and culture. This chapter identifies PhD student training as a curious process in which instruction and the advancement of science go together. It examines how a PhD student was instructed to tackle a common, though often challenging, problem of science with large datasets: calibrating a new dataset and combining it with data from a different source for analysis. By following this student around over two years as she achieved this goal, the author learnt how she became a competent member in the community and culture of extragalactic astronomy. Conversely, it is possible to gain insights into what makes combining scientific datasets often so challenging. As such, this chapter applies the tactics of Chapter 2 – take a problem of data-intensive science, consider how it is “staffed” in a specific case, and follow its management ethnographically – to another setting. This account serves as a resource for the next two chapters, on uses of diagrams and mundane reasoning in research with large datasets.

This chapter focuses on evaluations of persons as a lens into scientific data production and its ethics. That humans are fundamentally evaluative is a basic tenet of social interaction and social life. People are concerned that others understand their intentions adequately, knowing that their actions are being evaluated, and they examine others’ actions and intentions likewise. In many sciences, data production has become a service, with technicians generating data in the absence of researchers. Such new arrangements of data production come with numerical and social accountabilities as well as ethical evaluations. Focusing on data production in astronomy, this chapter traces these through several contexts. Joining data-producing technicians and data-using scientists as an ethnographer reveals that both are themselves exploring the epistemic and social accountabilities they face. Their way of “doing ethnography” is an ordinary social competence. The chapter argues that such ethnographic practices support and enable scientific data production as a service while also revealing its ethical tensions.