In inference control
|“||[s]ome cell values (eg small ones such as 1-5) in statistical data can present a greater risk of re-identification. Depending on the circumstances, small numbers can either be suppressed, or the values manipulated. . . . If a large number of cells are affected, the level of aggregation could be changed. For example, the data could be linked to wider geographical areas or age-bands could be widened.||”|
- ↑ Anonymisation: Managing Data Protection Risk Code of Practice, App. 2, at 52.