Skip to content

How to reduce bias in hiring interviews

By · Published · 4 min read

Hand-drawn illustration for the guide: How to reduce bias in hiring interviews

The most reliable way to reduce bias in interviews is to change the process, not to rely on people noticing their own biases. Ask every candidate the same job-related questions, score each answer against a written guide, have interviewers score alone before discussing, and decide what the job needs before you meet anyone. Awareness training can help people understand bias, but on its own it rarely changes hiring decisions. Structure does.

Common interview biases

BiasWhat happensExample
First impressionsAn early judgement colours everything afterA nervous first minute leads to a low score overall
ConfirmationInterviewers look for evidence that supports their first viewProbing hard on weaknesses only for candidates they doubt
SimilarityPreferring people like ourselvesWarming to a candidate from the same university
Halo and hornsOne strong or weak trait affects all ratingsGreat presentation skills raise the score for analysis too
ContrastJudging a candidate against the previous oneAn average candidate seems strong after a weak one
StereotypesAssumptions based on group membershipAssuming a parent cannot travel

Why awareness alone is not enough

Most biases work quickly and without our noticing. Knowing they exist does not stop them in the moment. Process changes work because they reduce the room for bias to act: everyone gets the same questions, answers are compared with the same standard and scores are recorded before group discussion.

Ten steps that reduce bias

  1. Define the job first. Agree the four to six skills that matter before you see any CVs.
  2. Write the questions in advance. Same questions, same order, for every candidate.
  3. Write scoring guides. Describe weak, average and strong answers for each question.
  4. Remove unnecessary information from CVs at screening where you can, such as photos, dates of birth and addresses.
  5. Use a work sample where possible: a short task like the real job.
  6. Use more than one interviewer and, where you can, a mix of backgrounds.
  7. Take notes on evidence, what the candidate said or did, during the interview.
  8. Score straight after each interview, alone, before talking to anyone.
  9. Discuss skill by skill in the debrief, with the most senior person speaking last.
  10. Review your data. Look at who applies, who is interviewed and who is hired. Gaps can show where the process needs work.

Questions to avoid

Ask only what relates to the job. Questions about age, family plans, marital status, religion, health, ethnicity, sexual orientation and similar topics do not predict performance, and in the EU, discrimination on many of these grounds in hiring is unlawful. If you need to know whether someone can meet a job requirement, such as travel or shift work, ask everyone the same way: "The role needs travel two weeks a month. Is that something you can do?"

This is general information, not legal advice. Employment law differs between countries: check with an employment lawyer or HR adviser.

An example

A software company in Athens notices that its interview panels rate candidates very differently. It introduces written questions and scoring guides, asks each interviewer to submit scores before the debrief, and adds a 45-minute work sample. After two hiring rounds, interviewers' scores for the same candidate are much closer, and the team spends less time arguing in debriefs.

What the evidence says

That structured interviews reduce the effect of interviewer bias and are fairer across groups than unstructured ones is Strong. That bias awareness training alone changes hiring decisions is Weak: studies often show short-term changes in attitudes but little lasting change in behaviour. That removing identifying information at screening reduces bias is Mixed: results vary by setting. Work samples predicting performance well is Strong.

Frequently asked questions

How do you reduce bias in interviews? Use the same job-related questions for every candidate, score against written guides, have interviewers score alone and decide what the job needs before you start.

Does unconscious bias training work? On its own, rarely. It can raise awareness, but research shows little lasting change in decisions unless the process changes too.

What are examples of interview bias? First impressions, confirmation bias, preferring people similar to you, halo effects, contrast with the previous candidate and stereotypes.

Check yourself

  1. Why do process changes work better than awareness alone? They reduce the room for bias to act, regardless of whether people notice it.
  2. What is the contrast effect? Judging a candidate against the previous one rather than the standard.
  3. How should you ask about a job requirement like travel? The same way for every candidate, tied to the job.

Go deeper: free sample lesson

The Structured Interviewing course covers interview bias and the process changes that reduce it. Read the free sample lesson

Written by , founder of Dokimos. He is a senior product and UX designer based in Athens, and is studying for a BSc in Psychological Science.