Virtual Archiving for Public Opinion Polls

Abstract

The Odum Institute for Research in Social Science Data Archive at the University of North Carolina, and partners from the National Network of State Polls present progress on a two year demonstration project using the Dataverse Virtual Archiving technology [1]. The goal of the Virtual Archiving for Public Opinion Polls: A Demonstration Project aims to streamline the ingest process and increase timely submission to data archives. Bridging this gap between producers and archives will increase the overall submission rates and ultimately preserve many data sets that would otherwise be lost. Around the world researchers and scientists collect vast amounts of data, which often are not archived after the completion of the project or task. As researchers move on to new projects, past data they have collected are seldom documented and preserved [6]. Until the tools for data curation are integrated into the research lifecycle of data, we will continue to experience this problem [2]. This project seeks to provide a solution for this problem. Although the virtual archiving technology needed to bridge the gap between the data producers and archives already exists, the availability of this tool and its value needs to be communicated to the scholarly community. The technology we use for this demonstration can be applied to many disciplines and data types. In this demonstration, we use public opinion data producers because these data serve as a useful, readily recognized example that will be widely replicated. Public opinion survey data are the most prevalent single kind of social science data and usually what most scientists first encounter. The Odum Institute’s relationship with the various state polling agencies and the National Network of State Polls make it an ideal candidate to propose new and innovative changes in the data life cycle of public opinion polls. This projects builds on our previous work with the Dataverse Network (DVN), developed at Harvard University. The Odum Institute has been an active partner in the DVN development and has recommended system modifications to allow for the maximum flexibility in public opinion preservation, distribution, and analysis. The DVN provides the tools necessary to implement this change. In this project, we use the DVN technology to aid data producing agencies in the ingest, curation and preservation of public opinion data, election polls and reports. Automated ingest tools specifically designed for quantitative data are used to create metadata automatically on ingest. This is critical for two reasons. First, metadata are essential in making data accessible to the scholarly community beyond those who were involved in the data collection. Second, the creation of metadata — itself a technical field with its own set of tools and norms — is a specialty that lies beyond the expertise of most research teams. The project is creating customized Web interfaces or “virtual archives” for each of the participating data producers. These interfaces are created to allow the researchers to seamlessly upload their data. The data will be transparently archived, preserved and curated by an organization trained in the field of long-term preservation. Once in the Dataverse Network, the data can be discovered and accessed via the existing federated search capabilities of the DVN. We are designing new workflows and help train the participating data producers to use these new and efficient methods of data ingest.

Details

Creators

Jonathan Crabtree

Institutions

Date

Keywords

singapore; digital archives; alliances; federation; data management; social science data

Publication Type

paper

License

CC BY-SA 3.0 AT

Download

504521 bytes

View This Publication