The Odum Institute hosts and manages the University of North Carolina Dataverse. The Dataverse Project, developed at the Institute for Quantitative Social Science (IQSS) at Harvard University, is an open-source software solution for preserving, searching and discovering, archiving, publishing, and analyzing data. It also makes it easy for researchers to receive recognition for their data contributions by automatically generating citations with persistent identifiers included. The Dataverse Project also supports a worldwide community of institutions that have adopted the Dataverse as their data repository platform. Members of the community provide feedback to Dataverse developers on system functionality and features, contribute code for Dataverse software enhancements, and share case studies on various uses of the Dataverse.
Odum Institute Data Archive + UNC Dataverse
Archival Data Collections
As curators and stewards of the Institute’s archival dataset collections, the Odum Institute Data Archive performs key tasks for preserving, documenting, and providing access to data using Dataverse tools. The UNC Dataverse provides public access to these valuable data collections curated and archived by the Odum Institute Data Archive. Review the Collection Development Policy to learn more about the Odum Institute Data Archive collections or procedures for contributing data to the Archive.
Data Repository Service
The Odum Institute Data Archive makes the UNC Dataverse software available to researchers to support their data management and sharing activities. Archive staff also offers training and guidance on use of the UNC Dataverse. For researchers seeking direct data management and archiving assistance, the Data Archive offers tiered data management and curation services to support varying data archiving and sharing needs.
UNC Dataverse User Support
Preparing Data for Archiving and Sharing
The following guidelines are provided to assist researchers in preparing data files for archiving and sharing that meet standards for long-term data preservation and reuse.
The UNC Dataverse automatically generates tab-delimited preservation copies of data files for specified file types. This enables the Dataverse to provide rich data support that allows users to explore tabular data in the Dataverse interface and download data files in multiple formats. Therefore, it is suggested that depositors upload quantitative tabular data in the following supported formats:
|IBM SPSS||.por OR .sav||Versions 7 to 22|
|Stata||.dta||Versions 4 to 13|
|R||.RData||Versions 1 to 3|
|Microsoft Excel||.xlsx||.xls is not supported|
|Comma-separated values||.csv||Limited support|
|Geospatial data||.shp AND .shx AND .dbf AND .prj||All four files are the minimum required to enable visualization|
To ensure that supporting documentation such as README files, reports, codebooks, questionnaires, and instruments can be preserved over the long-term, it is suggested that depositors upload document files in the following formats:
|Adobe Portable Document Format||.pdf/ua OR .pdf/a OR .pdf|
Other File Types
While the UNC Dataverse accepts all file formats, data depositors should still consider the long-term sustainability of all file formats. Software dependencies, proprietary status, limited adoption, lack of public documentation, and other factors can negatively impact the ability to render files in the future. Refer to the Library of Congress Recommended Formats Statement when selecting file formats for long-term preservation.
In order to ensure data remain usable into the future, it is important to provide accompanying documentation to provide enough context to allow for appropriate interpretation and use of the data. This documentation may include README files, methodology reports, codebooks/data dictionaries, lab protocols, and instruments. Data depositors should consider what supporting information is required to understand the content of data files as well as pertinent information about the creation, manipulation, and structure of the data.
The UNC Dataverse supports metadata standards for several disciplinary domains. Metadata describes data in a structured, machine-readable form that enables and enhances data discovery, interpretability, and reuse. The UNC Dataverse requires a minimum set of descriptive metadata based on the DataCite Metadata Schema standard, which facilitates accurate identification of the data for discovery and citation purposes. Additional domain-specific metadata blocks are available to allow data depositors to provide rich description to facilitate independent and informed use of the data. Data depositors are strongly encouraged to provide as much metadata as possible when submitting datasets to the UNC Dataverse.