The Limits of sharing: Controlled data collections

Abstract

We investigated 24 web-based data repositories with "controlled collections" to determine why and how repositories control access to and use of data. We selected our sample of data repositories from across scholarly and scientific disciplines in order to investigate differences between fields. Using content analysis and surveys, we collected data about current repository policies and practices and underlying motivations for controlling data access and use. Looking across all disciplines, we found no overarching reason for restricting access to data, but "avoiding misuse" was listed most frequently. Ensuring attribution was the dominant reason for controlling use of data. Observed between-field findings are tentative given the small number of repositories in some fields that met study criteria; however, our data do suggest some interesting differences. We also found cross-disciplinary patterns regarding methods for controlling access to and use of data. Better understanding of and attention to access and use control interests may allow repositories to attract more data depositors and ultimately increase the amount of data that can be shared.

Data sharing for the advancement of science: Overcoming barriers for citizen scientists

“…All of these reasons for controlling access to data expressed by the volunteers were also expressed by professional researcher participants in Eschenfelder and Johnson's () study. The importance of intellectual property and ownership are emphasized by Borgman () and Pryor ().…”

What Drives Academic Data Sharing?

“…Eschenfelder and Johnson [24] suggest more control for researchers over deposited data (see also [25][26][27][28][29][30][31][32]). According to some scholars, a priority right for publications, for example an embargo on data (e.g., [33]), would enable academic data sharing.…”

Environmental voluntary groups: Towards curating data for sharing, access and preservation

“…Furthermore, they were aware that sensitive data (such as the location of rare and endangered species) need protecting, and were not sure how the major repositories would manage this. The problem of possible misuse of data has been recognized in the literature (e.g., by Eschenfelder & Johnson, 2011).…”

“…The lack of acknowledgement and attribution of their data, which past contributors believed they had experienced, was also a deterrent to data sharing through large repositories. Issues connected to intellectual capital, ownership and intellectual property, and attribution are also discussed in the literature, in a broader context, (Borgman, 2006; Eschenfelder and Johnson, 2011; Pryor, 2009; Tenopir et al, 2011). A major consideration for sharing volunteer data, e.g., of individual observations or small datasets incorporated into larger ones, is that the individual sharer loses control of how the data are used, and who may use them once they are deposited, even if the individual datum is attributed.…”

“…Researchers view their data as intellectual capital and report a sense of ownership and responsibility (Pryor, 2009). As Borgman (2006, p. 360) stated; data sharing “is a complex social process involving trust, incentives, disincentives, risks, and intellectual property.” A study of controlled‐access data repositories finds many reasons for controlling access to data, such as perceptions that data may be misused and that sensitive data needs protecting, along with concerns about intellectual property and attribution (Eschenfelder & Johnson, 2011). These reasons may also resonate with non‐sharers of data as a recent study of data sharing practices by scientists found.…”