The Backstage Work of Data Sharing

Abstract

Conventional wisdom suggests that there are benefits to the creation of shared repositories of scientific data. Funding agencies require that the data from sponsored projects be shared publicly, but individual researchers often see little personal benefit to offset the work of creating easily sharable data. These conflicting forces have led to the emergence of a new role to support researchers: data managers. This paper identifies key differences between the sociotechnical context of data managers and other "human infrastructure" roles articulated previously in Computer Supported Cooperative Work (CSCW) literature and summarizes the challenges that data managers face when accepting data for archival and reuse. While data managers' work is critical for advancing science and science policy, their work is often invisible and under-appreciated since it takes place behind the scenes.

Ecological data sharing

“…First and foremost, they jealously value their time and have real concerns about the requisite time, labor and expertise to share data (Campbell et al, 2002;Tenopir et al, 2011). Researchers are also concerned about the potential for misinterpretation and misuse of data (Campbell et al, 2002;Davis et al, 2001;Hilgartner, 1997;Hilgartner and Brandt-Rauf, 1994;Kervin et al, 2014). Nevertheless, recent surveys indicate that most environmental and ecological scientists are willing to share their data, but they are challenged by a lack of experience with data management and insufficient training, a paucity of effective and easy-to-use metadata management tools, lack of awareness of standards, and absence of institutional support and resources for data management (Kervin et al, 2014;Tenopir et al, 2011).…”

“…Researchers are also concerned about the potential for misinterpretation and misuse of data (Campbell et al, 2002;Davis et al, 2001;Hilgartner, 1997;Hilgartner and Brandt-Rauf, 1994;Kervin et al, 2014). Nevertheless, recent surveys indicate that most environmental and ecological scientists are willing to share their data, but they are challenged by a lack of experience with data management and insufficient training, a paucity of effective and easy-to-use metadata management tools, lack of awareness of standards, and absence of institutional support and resources for data management (Kervin et al, 2014;Tenopir et al, 2011). Furthermore, numerous real and perceived legal constraints to sharing data exist such as different governmental and international approaches to copyright, the complexity of intellectual property rights and confidentiality issues, and uncertainty about the law (NSB, 2012;Reichman and Uhlir, 2003;Uhlir and Schröder, 2007).…”

“…data and metadata) published in the Ecological Society of America's Data Papers from 2004-2012 indicated that most authors did not provide metadata that was sufficient to support interpretation and re-use of the data (Kervin et al, 2013). Similarly, a recent survey of managers of environmental and ecological data repositories demonstrated that data contributors frequently made errors with respect to how data were organized (83% of the time) and documented (79% of the time) (Kervin et al, 2014).…”

“…Based on the lessons learned by examining peer-reviews of data publications (Kervin et al, 2013) and surveying repository managers (Kervin et al, 2014) and environmental scientists and ecologists (Tenopir et al, 2011), it is clear that data sharing and re-use can best be promulgated if several simple best practices are followed.…”

Dataset Reuse: Toward Translating Principles to Practice

“…), link to publication or project 3 , 13 , 54 , 60 , 63 , 66 Units and reference systems (1) defined, (2) consistently used 54 , 67 Representativeness/Population in relation to a total population 21 , 60 Caveats changes: classification/seasonal or special event/sample size/coverage/rounding 48 , 54 Cleaning/pre-processing (1) cleaning choices described, (2) are the raw data available? 3 , 13 , 21 , 68 Biases/limitations different types of bias (i.e., sampling bias) 21 , 49 , 69 Data management (1) mode of storage, (2) duration of storage 3 , 70 , 71 Documentation: Quality Missing values/null values (1) defined what they mean, (2) ratio of empty cells W3C 22 , 48 , 49 , 59 , 60 Margin of error/reliability/quality control procedures (1) confidence intervals, (2) estimates versus actual measurements 54 , 65 Formatting (1) consistent data type per column, (2) consistent date format W3C 41 …”