Data practices
The work of generating, managing and sharing data is an aspect of research that has received considerable attention in LIS and cognate areas, especially as academic libraries become more involved in curating and storing digital data for their constituencies. Librarians will increasingly take on responsibilities for the collection of both primary and secondary data and the processing, preservation and archiving required for sharing and reuse (Palmer & Cragin, 2008). In the sciences, research is becoming increasingly data intensive, and there are growing expectations that data resources will be aggregated and shared within and across disciplines. The collective management of data, however, does not fit with existing library models for managing other scholarly communication resources at the local or more global levels (Cragin & Shankar, 2006). Libraries have established infrastructure for acquiring, maintaining and providing access to many types of published materials, but they are not designed to accommodate the very different structures and uses of datasets.

Meeting data curation responsibilities will require a deep understanding of how researchers presently work with their data and of the potential of various kinds of data for future research. As with other types of information work, data practices are influenced by researchers’ disciplines and subdisciplines and other organizational and collaborative arrangements. While it is apparent that varying practices need to be taken into account in the development of digital data collections, and the computing networks and information infrastructures in which they reside, the specific functions and roles of data as information resources are not yet well understood (Bowker, 2000; Hine, 2005).

At the same time, the growth of digital data is clearly having a transformative effect on many Science. For example, in experimental neuroscience the function of standard brain atlases has been extended as an organizational structure for bringing together digital materials on regions of the brain, and visualization of brain imaging data has fostered important new analytical approaches (Beaulieu, 2004). Data repositories have been developed in fields where centralized data have been considered fundamental to the advancement of science, as in the case of GenBank and the Protein Data Bank (CM Brown, 2003). In some sciences it has become common practice to submit data as part of the peer review publishing process. While such practices are not yet widespread, concerns about managing and preserving digital data are growing across fields. A large-scale survey in the UK showed that computerized datasets were considered

Articles, Library management, Research