Outputs

Whether using safeguarded data or controlled data, Statistical Disclosure Control (SDC) techniques should always be applied to the work (“outputs”) produced.  This is especially important for controlled data due to the greater degree level of detail and concomitant higher risk of re-identification (and attribution).

Controlled data is available only through the Secure Lab or similar environments. Researchers must conduct all of their analysis, and produce outputs (such as papers, presentations etc.) within this area. Controlled data will not be released to be used in any other environment under any circumstance. Outputs will be returned to users following a full – and successful - SDC examination by two trained members of Service staff.

An SDC examination entails checking that all rules and principles relevant to these data have been adhered to, such as minimum cell counts, secondary disclosure, dominance and more.  These rules are not designed to inconvenience researchers, but to help satisfy the requirements of the legislation.  Good SDC goes hand in hand with good research practice.

Users of controlled data must attend a training course, and should maintain familiarity with SDC principles and practice. Whilst it deals primarily with data access in safe settings (i.e. controlled data, such as is found in the UK Data Service Secure Lab and similar environments), the statistical disclosure control principles outlined in the SDC Handbook, published by the Safe Data Access Professionals (SDAP) group, can be usefully applied by users of Special Licence data. Thinking about statistical disclosure control, whether using controlled or safeguarded data is simply good practice.  There are versions of the SDC Handbook in both English and Spanish.

If researchers are unsure about SDC when they produce outputs, we recommend that they speak to an User Support and Training Officer.