Methods
for Reducing Disclosure Risks When Sharing Data:
Overview of
Technological Methods
Rather than release data
altered to protect confidentiality, data disseminators can restrict
who gets access to the unaltered data. Secure data enclaves
make data available only to approved individuals in tightly
controlled locations. Remote access allows researchers to
access confidential data by connecting to a remote computer, but the
data cannot be saved on the researchers' machines. Remote execution
allows researchers to submit requests for output from statistical
analyses to a remote computer, which runs the analysis and reports
results without ever letting the analyst see the data. Data licensing
makes data available only for those who agree to terms set by the
data provider. Licensed individuals generally can save the data on
their own machines, although typically under rules about further
sharing. Excellent summaries of the pros and cons of these approaches
are in the National Academy of Science's 2005 report, Expanding
Access to Research Data,
and the Confidentiality and Data Access Committee's report
on restricted data access.
Examples of these approaches are listed below.
1.
Secure Data Enclave
2. Remote Access
3. Licensing