The project aimed at testing disclosure control methods that might be suitable for the protection of Census data in the ESS for the Census 2021 round, suggesting statistical disclosure control solutions for census hypercubes and grid data, and inviting Member States to experiment with alternative ways to do the confidentiality treatment, eventually leading to a more harmonised approach and less suppressed cells.
On this page, testers from other Member States will find four files provided for testing.
The SAS software codes supplied for testing in the project:
- Record swapping,
- Random noise (cell key method),
as well as a short note with instructions.
(See Part I of Deliverable 3.1 which gives an overview of the methodology and Part II for a more detailed description of the codes provided.)
When applying a disclosure control method, testers are also asked to compute the information loss measures suggested for the project (see deliverable 3.1., part II, section 5). Since the project aims for comparable results, testers will find an exemplary SAS code for computing the measures (prepared for hypercube 9.1) and an Excel file that should be used to document the test results.
RELATED DOCUMENTS
- Instructions
- Record Swapping
- Random Noise (cell key method)
- Information loss measure and test results - The SAS code is only provided for HC91 and it needs to be adjusted when applied to other hypercubes or grid data.