aggregation

Combining several datasets into one, which can make the result more sensitive than any of its parts on their own.

In more detail

A list of staff names and a table of salary bands may each be internal, but joined together they become payroll. Combined data can also re-identify people in datasets that looked anonymous. When combining data, start from the highest label among the inputs and then ask whether the combination itself deserves a higher one.

Example

Staff names (internal) joined with salary bands (internal) = payroll (confidential or higher)

Related terms

Taught in

References