Research Using Publicly Available Data Sets: UM Policy, revised May 2008
A common research method involves secondary analysis of publicly available survey data. The federal government provides public access to several important data sets (e.g., U.S. Bureau of the Census), and many federal funding programs now require that researchers make the data they collect publicly available. Likewise, many professional organizations and journals have a standard requirement that research data sets of published works be made accessible to encourage scholarly replication of research.
Under the federal regulations for human subjects research (45 CFR Part 46) publicly available data sets that are stripped of identifiers do not require IRB review. Because it may be difficult to understand the definition of "publicly available" and also, what "stripped of identifiers" means, upon recommendation of the IRBs, the university has instituted the following policy for research projects involving certain data sets:
Policy for Use of Publicly Available Data SetsResearch projects involving analysis of secondary data from any one of the following data sets/repositories will NOT require prior IRB approval, unless the archive hosting the data explicitly requires prior IRB approval before releasing the data for use. Note: If you are designing a research project that merges more than one public data set and you recognize that this may increase the risk of identification of individual research participants, then you should consult the IRB.
- Inter-University Consortium for Political and Social Research (ICPSR)
- National Center for Health Statistics
- National Center for Education Statistics
- National Election Studies
- Roper Center for Public Opinion Research
- The University of Michigan Health and Retirement Study (HRS)
- Unrestricted data sets only
- U.S. Bureau of the Census
- Panel Study of Income Dynamics (PSID)
- Survey of Consumers (SCA)
- Integrated Public Use Microdata Samples - International (IPUMS-i)
- Demographic and Health Surveys (DHS)
- (including the 2003 Demographic Health Survey of Nigeria)
This policy was first approved by the IRB Council in April 2006 and is revised as new data sets are approved by the Council.
Submitting a Data Set for Pre-approval
Data sets that may quality for inclusion on UM's list of approved data sources include:
- Public use data sets posted on the Internet that include a responsible use statement or other confidentiality agreement for authors to protect human subjects (for an example, see the ICPSR's confidentiality review: http://www.icpsr.umich.edu/access/deposit/conf-review.html)
- Survey data distributed by UM principal investigators who can
- the data collection procedures were approved by an IRB that satisfies the Common Rule criteria for an IRB, and
- the data set and documentation as distributed do not contain information that could be used to identify individual research participants.
- Name of data set
- URL of the data set or other information on how to obtain the data set
- Abstract (one page maximum) describing the content of the data set and its potential use.
If the subcommittee approves the request, the data set will be added to the list.