Practices Do Not Make Perfect:
Disciplinary Data Sharing and Reuse Practices and Their Implications for Repository Data Curation

As sharing and reusing research data have become commonplace, library, museum, archive, and data repository staff have had to evolve and meet new needs. It is challenging, given the varied data sharing and reuse practices and traditions at play within and across the different research communities.

In this book chapter, we discuss the differences among the social science, archaeological, and zoological communities.  Based on interviews and observations with 105 researchers reusing data within these disciplines, we also report 1) how they develop trust in the data they reuse and 2) what sources they engage with to gather contextual information about the data.

Download chapter .pdf

Full book .pdf   

Chapter originally published in: Curating Research Data, Volume 1: Practical Strategies for your Digital Repository. Lisa R. Johnson, Ed. Chicago: Association of College and Research Libraries.

The top 5 trust markers used when deciding whether to reuse data were:

1. Data producer reputation
2. Documentation quality
3. Original peer review publications about the data
4. Prior reuse of the data
5. Repository reputation

The top 7 sources engaged with to gather information about the data were:

1. Peer-reviewed publications
2. Repository/museum records
3. Data producer records
4. People
5. Documentation
6. Codebooks
7. Specimens/artifacts

Highlights:

• Researchers’ trust in data is informed by more than their encounters with data. Perceptions and opinions about documentation quality, data producer reputation, and repository reputation are formed over time as they gain experience with the discipline, data, and repository. Repository staff have to power to shape researchers' perceptions and opinions about some of these trust markers, particularly repository reputation and documentation quality. These markers provide a foundation of trust in the data. They may not be the only factor, but they are starting point in many cases.  Don’t make researchers guess.  Tell them directly, explicitly.
• Data repositories do not have to house all of the information about the data to be effective; but they do have to provide provenance and links to the information if it is housed elsewhere. We see this within the zoological community in particular where informed decisions about data stewardship and data services were made and partnerships were formed. Repository staff interested in considering similar actions should consider forming partnerships with other institutions to complement and extend each institution’s capabilities and to add value to the designated community of users.
• Despite the growth in repositories and how well established they have become, people remain an important source of information about the data, particularly data producers. By mediating information exchanges between data producers and reusers, repository staff could benefit with improved understanding of unmet needs and adapt accordingly. It might lead to user interface design changes, instructional modules or other scaffolding to improve data reusers’ experiences.
• Researchers across the disciplines engage with peer-reviewed publications to access data and contextual information or gauge their communities’ interest and acceptance of data for reuse. In disciplines where data sharing and reuse are new and peer-reviewed publications of data reuse studies are limited, repository staff should consider assembling a team to do peer review of the data, capturing data reuse metrics, using DOIs and suggesting a standard form for data citations, and maintaining bibliographies of data reuse studies.

Book also available from the ALA website.