What Does Peer Review of Data Sets Mean and What Roles do Data Archiving and Quality Control Have in the Processes?

At national, international, and local levels, data sets have always been essential for science research, but now need to be more visible as first-class scholarly objects. Many initiatives are establishing procedures to publish and curate data sets, as well as promoting scholarly reward for researchers that collect, create, manage, and preserve data sets. Traditionally, research quality has been assessed by peer review of textual publications, e.g. journal articles, conference proceedings, and books. Citation indices then provide standard measures of productivity used to reward individuals for their peer-reviewed work. Whether a similar peer review process is appropriate for assessing and ensuring the quality of data sets remains as an open question.

How does the traditional process of peer review apply to data sets? In this poster, we will present current work being done at the National Center for Atmospheric Research (NCAR) in the context of the Peer REview for Publication & Accreditation of Research Data in the Earth sciences (PREPARDE) project. PREPARDE is assessing practices and processes for data peer review, with the goal of developing recommendations. NCAR data management teams perform various kinds of quality assessment and review of data sets prior to making them publicly available. The poster will investigate how notions of peer review relate to the types of data review already in place at NCAR. We highlight the data set characteristics and management/archiving processes that challenge the traditional peer review processes by using a number of questions as probes, including: Who is qualified to review data sets? What formal and informal documentation is necessary to allow someone outside of a research team to review a data set? What data set review can be done pre-publication, and what must be done post-publication?