Redundancy Selection Record De-Identification with anonym.plus

Strip names from a selection matrix while the scores and dates a fairness review needs remain.

In simple terms, PII redaction is the on-device process of finding and masking personally identifiable information in a document before it is shared.

Redundancy selection de-identification is the removal of names from a scoring matrix, leaving the columns a fairness review depends on. Age is a protected characteristic under Equality Act 2010 s.5, so a selection that bites harder on older staff can be direct discrimination under s.13 or indirect under s.19. anonym.plus marks the names on your device, so each figure and score stays while the people are shielded.

When this applies

A matrix of scores, job titles, and dates of birth is evidence for two tests at once. Under s.19 you compare outcomes across age bands, which needs the dates. Under ERA 1996 s.98(4) a tribunal asks whether selection fell within the band of reasonable responses, which needs the scores. The statutory redundancy payment in ERA 1996 s.162 is itself age-banded — half a week's pay per year under 22, one week from 22 to 40, one and a half weeks at 41 and over — so age data cannot simply be dropped.

How anonym.plus handles it

  1. Open the matrix in anonym.plus on your device.
  2. The tool flags names while leaving the scoring columns.
  3. Local OCR reads a scanned spreadsheet page.
  4. Keep the figures and titles the fairness review uses.
  5. Swap each name for a steady label.
  6. Save the clean matrix locally.

What you need to provide

PII entity types detected

Categoryanonym.plus entity typeExample
NamesPERSONpool member → [WORKER]
DatesDATE_TIMEage 54 → 54 (kept)
NRPNRPjob title → [TITLE]
IdentifiersNATIONAL_IDstaff no. 90213 → [STAFF_ID]
OrganisationORGANIZATIONdepartment → [DEPT]
ContactEMAIL_ADDRESSwork email → [EMAIL]

Compliance achieved

Anonymise redundancy selection records offline — see plans & start free →

Limitations & cautions

A fairness review needs the figures and scores, so removing them would defeat the check. Yet a rare title plus a date of birth can re-identify one person in a small pool. Band or group small cells where the review allows, and keep the unredacted matrix on your own file.

Frequently asked questions

Why keep dates of birth on a selection matrix?

An age discrimination check under Equality Act 2010 s.13 and s.19 compares outcomes across age bands, and ERA 1996 s.162 calculates the statutory payment from age and service. Drop the dates and neither works.

Can a date of birth plus a title still identify someone?

Yes, in a small pool. Where the review permits, band the ages or group rare titles so one row cannot name a worker.

Is the matrix processed in the cloud?

No. The whole run is offline, so the data stays on your device.