Skip to content

Anonymity threshold

Also called minimum group size, minimum reporting size, small cell suppression

The smallest number of people a result may describe. Any summary of fewer is hidden, so nobody can be picked out of it.

An anonymous survey can still identify people through its results. If a filter or a breakdown narrows the answers to two people in one team, anyone who knows the team can often tell who said what. A threshold, commonly five, stops any result from describing fewer people than that.

It has to hold everywhere a result can be narrowed: the whole survey before enough have answered, a filter to one group, each row and column of a cross-tab, a comparison of one segment with everyone else, and every report shared outside the team. Totals need care too. A table that hides a group of two but still prints the full total gives the hidden group away by subtraction.

Small groups can be merged into one larger group instead of being dropped, which keeps their answers in the picture without singling anyone out.

Where it goes wrong

Thresholds that hide the summaries but not the rows. Whoever can open individual responses can still read them, so say plainly who that is.

See also

De-identification
Removing or replacing the details that could identify a respondent, so what you hold cannot be traced back to a person.
Cross-tabulation
A table that breaks down answers to one question by answers to another.
Net Promoter Score
The share of people scoring 9 or 10 on a recommend question minus the share scoring 0 to 6, in percentage points.

All 35 terms

Start

Put it into practice in your next survey.