Skip to main content

Research data across the lifecycle

An overview of considerations, resources and tools for working with data in your research projects

Why share data?

There are several possible reasons for wanting or needing to share - make globally and openly accessible, or with some restrictions if necessary - research data:

  • To enable replicability and openness of research - for the benefit of science, other researchers, the public, policymakers, journalists, students, ...
  • To potentially increase citations of the publication(s) based on the research data  (see: Sharing Detailed Research Data Is Associated with Increased Citation Rate). Researchers finding research data that is linked to the article or other empirical work based on that data may increase the exposure, reading, and citation of that work.
  • To potentially increase the impact of one's research by making its data available to a broader audience than would read, or could even access, the publication(s) based on the data.
  • A research funders or publisher requires it.

One way for AU faculty to share data is to submit it to the AU Digital Research Archive.

 

Submitting data to the AU Digital Research Archive

AU researchers can specify the AU Digital Research Archive (AUDRA.american.edu), which has a collection for faculty-produced research data, as a vehicle for sharing and archiving their data in a data management plan.  The AUDRA, launched in the summer of 2015, is based on the Islandora platform, and hosted and operated for AU by the IT group of the Washington Research Library Consortium (WRLC).  The current dataset size limit is 2½ GB per file for uploads through AUDRA's web interface; WRLC can import larger files sent to them on disk (contact me to discuss details if that need arises).

Since many data management plans require addressing which metadata standard(s) of the researcher's community will be employed in data sharing, it is noteworthy that the AUDRA uses the Metadata Object Description Schema (MODS) for all digital objects it hosts. These metadata elements, then, are part of the AUDRA platform; additionally, researchers may submit, along with datasets, files containing metadata for them (for example, DDI XML files created with any of the available DDI Tools). See also: Disciplinary Metadata. Furthermore, data (and related research outputs) submitted to AUDRA can be assigned globally unique, persistent Digital Object Identifiers (DOIs); see Digital Object Identifiers (DOIs) and their use at American U.

‚ÄčAU researchers can submit datasets to AUDRA using the Faculty Research Deposit form. For submitted file formats, they should consider whether these are suitable for long-term usability and software-independence, as opposed to being in a proprietary format that is only readable with specific software (which may become unavailable or change what file formats it supports).  For guidance, see, for example: Sustainability of Digital Formats: Datasets: Quality and Functionality Factors and Format Descriptions for Dataset Formats by the Library of Congress. (AU library personnel will generally not modify or convert researcher-submitted datasets.)  See also tools for working with research data.

Files submitted to AUDRA can, upon the submitter's request, have an embargo date applied to them, so that they do not actually become publicly available until that future date.