Two interviewers can conduct the same interview and yet come away with different impressions. Usually, this is not down to the candidate, but to the way our brains select, interpret and remember information. Who reducing bias in job interviews If you want to take this seriously, you must first learn to recognise where things go wrong and then organise the process in such a way that you leave less room for arbitrariness.
What is interview bias (and why is it so persistent)?
Interview bias is any systematic distortion that causes your assessment to differ from what someone is likely to demonstrate in the role. It is therefore not a matter of “a one-off misjudgement”, but of patterns that recur predictably.
Bias often arises from time pressure, scarcity (“we have to find someone”), an interview structure that is too loose, or too much emphasis on first impressions. The brain wants to make sense of things quickly: does this person fit my image, yes or no?
The tricky thing is that bias often masquerades as a rational conclusion. Phrases such as “I just don’t see it” or “this doesn’t feel like a senior role” sound well-founded, but are often simply a summary of impressions without any evidence.
Why this also poses a risk to your organisation
A subjective interview not only leads to poorer hires, but also to decisions that are difficult to justify. This becomes particularly tricky if a rejected candidate asks for the reasoning behind the decision, or if there is an internal debate about ‘who was the best’.
In the Netherlands, this also relates to equal treatment. The legislation is clear on the prohibition of discrimination in employment, including recruitment and selection. See the explanation of the Central Government on discrimination for the legal basis.
The most common forms of bias in interviews
Bias is rarely just one thing. In practice, small distortions accumulate, which can skew the final decision.
Would you like to find out more about data-driven recruitment?
- Recruitment using AI and data
- Takes an average of 3–8 weeks to complete
1) Halo and horn effects
A single striking positive aspect (halo) or a single negative aspect (horn) colours your judgement of everything that follows. A candidate who makes a strong first impression is given more of the “benefit of the doubt”; someone who starts off nervously suddenly has to work harder to earn the same level of trust.
- Signal: you use broad labels such as “professional” or “clumsy” without any concrete evidence.
- Risk: you’re less likely to notice behaviour that doesn’t fit with your initial impression, or you’ll forget it later.
2) Similarity bias (liking and recognition)
We often find people more likeable when they are similar to us: the same background, sense of humour, communication style or working style. This can feel useful (“they’ll fit into the team”), but it’s a poor predictor of performance if you don’t translate it into job-related behaviour.
- Signal: the arguments are about chemistry, the vibe or “I’d like to go out with him”.
- Solution: Define what you mean by ‘cooperation’ or ‘communication’ in terms of observable behaviour.
3) Confirmation bias
As soon as you have a hypothesis (“too junior”, “job-hopping too much”), you unconsciously start asking questions that confirm it. Contrary evidence is given less weight or is dismissed.
This happens particularly often if you’ve already discussed CVs at length before the interview. In that case, you’ll end up using the interview to confirm an opinion that’s already all but set in stone.
4) Anchoring (the first number sticks in your mind)
The first impression, or the initial assessment, acts as an anchor. Everything that comes after is shaped around it. You can see this happening when the first interviewer is very assertive and the rest unconsciously follow suit.
- Signal: everyone ends up scoring “more or less the same” without any real evidence of their own.
- Solution: Let interviewers score the candidates independently first, and only then discuss the results.
5) Affinity bias regarding school, language and ‘neatness’
In many roles, a preference for certain educational backgrounds, accent-free Dutch or a polished manner of presentation creeps in. Sometimes this is relevant (for example, in customer-facing roles), but often it is simply a proxy for “familiarity”.
A practical check: can you identify which behaviours in the workplace break down when someone is less talkative, or is it mainly a matter of discomfort in the interview room?
Reducing bias in job interviews through process choices
You don’t need to “eliminate all bias”. That’s not possible. However, you can reduce the scope within which bias determines the outcome by making a few choices that work consistently.
Use a structured interview
A set list of questions and a fixed order force you to gather comparable evidence. It prevents you from probing one candidate in depth about their achievements whilst mainly discussing hobbies with another.
If you want to put this into practice, it ties in with How to conduct a structured interview in 8 steps. This approach helps to establish in advance what you are measuring and how you will assess it.
Use a scorecard with behavioural anchors
With just “1–5”, you’re still scoring based on gut feeling. Behavioural anchors make it measurable: what constitutes a 3 and what constitutes a 5, and what evidence goes with each?
A practical starting point is Create a selection scorecard for hiring managers, so that you aren’t judged by different standards depending on the interviewer.
Introduce ‘interview etiquette’ as a team agreement
There’s a lot to be gained from small routines that prevent decisions from being made too soon.
- Record facts and quotations, not interpretations (“X said”, not “was uncertain”).
- Give marks immediately after each answer, not just at the end as a reminder.
- Please save the discussion about “fit” until after the core competence scores have been announced.
- Only compare candidates once you have completed at least two interviews.
Calibration: the same answer, the same score
Plan a single short session (30–45 mins) in which interviewers jointly score sample answers. If one interviewer gives a 2 and the other a 4, the problem lies not with the candidate but with the interpretation.
Calibration doesn’t make your selection overly rigid; on the contrary, it saves you from having to discuss it afterwards.
Work with several assessors, but with clearly defined roles
Having more interviewers can reduce bias, but only if everyone looks at different information. Don’t have three people asking the same general questions.
- Interviewer A: core competencies (STAR) and evidence from previous situations.
- Interviewer B: case study or practical exercise and thought process.
- Interviewer C: preconditions and context (shifts, journey time, safety requirements).
A simple table for identifying bias risks
This table helps you to quickly identify, during your debrief, where the conversation may have gone off track.
| Feedback during the debriefing | Probable bias | What you’re asking now |
|---|---|---|
| “I just don’t click with them.” | Similarity/affinity bias | What exactly does this entail in terms of behaviour at work? |
| “He’s so convincing, so this is going to be brilliant.” | Halo effect | What evidence of results have we seen? |
| “She’s probably too junior.” | Confirmation bias | Which answer or example illustrates this? |
| “Interviewer 1 gave a 5 straight away.” | Anchoring | What were our scores for each criterion, assessed independently? |
Things you should avoid doing if you want to make fairer selections
Some interventions seem logical, but actually make the conversation more subjective.
- Chatting casually “to get to know someone”: You’re mainly measuring social style and similarities.
- Too much emphasis on the CV: you’re selecting people based on their past, rather than on evidence of the behaviour you’re looking for.
- Holding a discussion without a scorecard: It is often the most assertive person who wins, not the best argument.
- Stop looking once you’ve found one good candidate: A sense of scarcity makes you more willing to take risks.
If you recognise these patterns in your current process, then this also ties in with the pitfalls outlined in These 4 mistakes are costing you your best candidates, particularly with regard to jumping to conclusions and the lack of objective methods.
Practical next step: make it measurable in a single job vacancy
Choose one role where there is often debate about “who was the best”. Set up a structured interview and a scorecard for that role, and after 4–6 interviews, assess the results: less confusion, quicker decisions, better justification.
If you’d like to discuss an approach suited to blue-collar, sales or maritime roles, FosFor can help with scorecards, interview structures and data-driven decisions throughout the recruitment funnel. A short session to review your process against these bias points is often enough to deliver quick results.