Signal S10

Network overlap with employees

Remove

The vendor says: “Following Kestrel pages or being connected to Kestrel employees indicates culture fit.”

Moves the score by up to +10 pointsFeeds: culture fit

The six questions

1. Is it related to the job?
Knowing current employees says nothing about ability.
Fail
2. Does it stand in for who someone is?
Tracks college tier (strong), family income and background (weak). Existing employees cluster in a few colleges, so their networks do too.
Fail
3. Is it the right person?
Connections belong to the account.
Pass
4. Did the candidate publish it, and would they expect an employer to use it this way?
Connections are made socially.
Concern
5. Does it work equally for everyone?
Candidates from colleges Kestrel has rarely hired from start behind.
Concern
6. Is there a less intrusive way?
No less intrusive way to learn the same thing was identified.
Pass

Rule applied

If a signal stands in strongly or directly for a personal characteristic (question 2 fails), remove it.

Conditions

None. A removed signal is not used under any condition.

A retail company's recruiting model learned to penalise women (2018)Evidence
Amazon abandoned an experimental résumé-scoring model after finding it downgraded résumés containing the word "women's" and graduates of two women's colleges. It had learned from ten years of mostly male applicants' résumés.
US Uniform Guidelines: the four-fifths ruleReference only
A selection rate for any group that is less than four-fifths (80%) of the rate for the group with the highest rate is generally regarded as evidence of adverse impact.
New York City Local Law 144: bias audits of hiring toolsReference only
Employers using automated employment decision tools must commission an annual independent bias audit reporting selection rates and impact ratios by group, publish a summary, and notify candidates in advance.

Aditya Malhotra · LinkedIn · English
“Grateful to the Kestrel alumni who mentored me through placement season.”