Research data and metadata curation as institutional issues

TLDR

Research data curation initiatives must support heterogeneous projects, data, and metadata, and institutions—stable patterns of human behavior—structure how these practices originate, stabilize, evolve, and transfer. This article examines variability in data and metadata practices using institutions as the key theoretical concept. The authors present a theoretical framework of five institutional carriers—norms and symbols, intermediaries, routines, standards, and material objects—and apply it to case studies of CENS, LTER, and UCAR. The case studies show that institutional support for data and metadata management varies widely within and across disciplines and organizations, revealing a broad spectrum of configurations.

Abstract

Research data curation initiatives must support heterogeneous kinds of projects, data, and metadata. This article examines variability in data and metadata practices using “institutions” as the key theoretical concept. Institutions, in the sense used here, are stable patterns of human behavior that structure, legitimize, or delegitimize actions, relationships, and understandings within particular situations. Based on prior conceptualizations of institutions, a theoretical framework is presented that outlines 5 categories of “institutional carriers” for data practices: (a) norms and symbols, (b) intermediaries, (c) routines, (d) standards, and (e) material objects. These institutional carriers are central to understanding how scientific data and metadata practices originate, stabilize, evolve, and transfer. This institutional framework is applied to 3 case studies: the C enter for E mbedded N etworked S ensing ( CENS ), the L ong T erm E cological R esearch ( LTER ) network, and the U niversity C orporation for A tmospheric R esearch ( UCAR ). These cases are used to illustrate how institutional support for data and metadata management are not uniform within a single organization or academic discipline. Instead, broad spectra of institutional configurations for managing data and metadata exist within and across disciplines and organizations.