November Highlight on Data Repositories: ICPSR

This month, we’d like to highlight the Inter-university Consortium for Political and Social Research (ICPSR): http://www.icpsr.umich.edu/

What is ICPSR?

ICPSR was founded in 1962 at the University of Michigan and now exists as a unit within the Institute for Social Research. It is the world’s largest social science data archive, with over 7,000 data collections and 500,000 individual data files that can be browsed by topic or searched. As of fall 2012, it has over 700 member institutions, including the University of Texas at Austin. Data is contributed by individual researchers, government agencies, and research organizations. ICPSR maintains a citation database of data-related literature to facilitate literature searches and the study of data as intellectual output. It is an international leader in data management and digital preservation dedicated to ensuring long-term usability of data.

How does deposit work?

Deposits must include all data and documentation necessary to read and interpret the data collection. For researchers interested in depositing their data, ICPSR’s Guide to Social Science Data Preparation and Archiving describes best practice for preparing data to be shared. ICPSR offers a secure electronic deposit form for researchers to upload and describe their data. More information about deposit is available here. Once data are submitted, data processors review data for confidentiality issues, convert documentation to electronic, PDF/A form, generate multiple data formats for dissemination and preservation, create Data Documentation Initiative-compliant documentation, create a descriptive metadata record, and assign the dataset a Digital Object Identifier. Once deposited, dataset usage can then be tracked through Utilization Reports.

What can be deposited?

Using the online secure deposit form, up to 2 GB can be uploaded. Preferred file formats are as follows:

  • Quantitative data files: SPSS, SAS, Stata
  • Qualitative data files: ASCII, RTF
  • Audio files: AIFF, WAV
  • Video files: MPEG4, JPEG2000
  • Documentation: ASCII, DDI-XML, Microsoft Word (PDF is acceptable)

How does ICPSR manage sensitive data and confidentiality?

ICPSR offers several deposit options for sensitive data.

For traditional restricted data, researchers interested in using the data must belong to membership institutions and fill out an application about their research. These requests will then be reviewed by ICPSR staff to ensure all security requirements have been fulfilled and the data will be sent via mail on a CD/DVD. For an additional layer of security, ICPSR can send information to an external body for review if necessary.

For highly sensitive data, data can be restricted to only on-site analysis at the University of Michigan’s Perry Building enclave. Investigators wishing to use materials restricted in this fashion must sign an Application for Use of the ICPSR Data Enclave and Confidentiality Agreement along with an official of their home institution. At the enclave, only the provided computer can be used and materials are reviewed for disclosure risk before leaving. All analysis output is evaluated by an ICPSR unit manager and sent to the researcher on a CD/DVD at after approval.

For simple analysis of sensitive data, ICPSR offers the Survey Documentation and Analysis statistical package that can evaluate output for disclosure risk before displaying it to the end user. More information is available here.

ICPSR will preserve data under a delayed dissemination model if necessary. They will preserve data until a predetermined release date and distribute normally after that date.

ICPSR is working on a virtual data enclave to permit remote access and analysis for sensitive data, which researchers cannot download, copy, or paste. The analysis output will then be evaluated by ICPSR staff before being released. This virtual data enclave is not yet operational.

More information on confidentiality is available here.

All of the information provided comes from http://www.icpsr.umich.edu/ and   Johnson, W. G. (2008). The ICPSR and social science research. Behavioral & Social Sciences Librarian, 140-157.  doi:10.1080/01639260802385200.

Posted by Franny Gaede on November 1, 2012 at 3:18 pm

No Comments