Definition · Interviewing and assessment
Interview Bias
Definition
Interview bias is any systematic error that leads an interviewer to rate a candidate on something other than their ability to do the job, such as a first impression, a shared hobby, or one standout answer. It is mostly unintentional, and it is reduced far more by process design than by awareness training alone.
Also called: Interviewer bias, Bias in interviews, Types of interview bias, Rater bias
Types of interview bias you will actually see
- First-impression bias. The handshake, outfit or first answer sets a verdict, and the rest of the interview is spent confirming it.
- [Halo effect](/glossary/halo-effect/). One strong trait, like a prestigious employer or polished speaking, raises ratings on unrelated competencies.
- [Horns effect](/glossary/horns-effect/). The reverse: one weakness, such as a nervous start or a typo on the resume, drags everything down.
- [Affinity bias](/glossary/affinity-bias/). Favoring candidates who share your school, hometown, interests or background.
- [Confirmation bias](/glossary/confirmation-bias/). Asking questions that confirm what you already believe from the resume, and discounting evidence against it.
- [Contrast effect](/glossary/contrast-effect/). Rating a candidate against the person just before them rather than against the standard. An average candidate looks great after a weak one.
- Central tendency and leniency. Rating everyone a 3, or everyone a 4, to avoid making a call.
- Recency bias. Remembering the end of the interview far better than the start.
- Nonverbal bias. Over-weighting eye contact, accent or body language, which often reflect culture or disability rather than job skill.
- Stereotyping. Assuming traits from group membership, for example that an older candidate will struggle with new software.
Why it matters beyond fairness
Biased interviews are less accurate interviews. Each bias adds noise that has nothing to do with job performance, which means you hire some people who struggle and reject some who would have excelled. In high-volume hiring that error repeats hundreds of times a year.
There is also legal exposure. Title VII, the ADEA and the ADA prohibit discrimination in hiring decisions, and a pattern of lower pass rates for a protected group at the interview stage can support a disparate impact claim even if nobody intended it. Interview notes and ratings are frequently reviewed in those cases.
Process changes that reduce bias
Telling people about bias helps a little. Changing the process helps much more:
- Use a [structured interview](/glossary/structured-interview/). Same questions, same order, same rating anchors for every candidate.
- Define what good looks like before you meet anyone. Write rating anchors in advance so the standard does not move to fit the candidate in front of you.
- Take notes on evidence, rate afterwards. Score each competency separately, not with one overall feeling.
- Score independently. Interviewers submit ratings before discussing or seeing each other's scores.
- Use multiple interviewers or a panel with mixed backgrounds, so individual quirks average out.
- Limit unstructured chat time. Small talk is where affinity bias thrives. Keep it brief and do not score it.
- Audit outcomes. Compare pass rates at each interview stage by gender, race and age band where data allows, and investigate gaps.
These steps overlap heavily with the general advice for reducing unconscious bias in hiring, but they are aimed at the interview specifically.
A quick example
Two interviewers at a logistics company's regional hub assess the same shift supervisor candidate. One spent five minutes bonding over a shared college football team before starting; she rates the candidate 5 on leadership. The other rates him 3. In the debrief, the evidence for leadership turns out to be one general statement ("I always lead by example") and no specific example of managing a performance problem. With written anchors that require a concrete example for a 4 or 5, the rating comes back to 3, and the debate is over in two minutes.
Bias in AI and automated interviews
Automated interviews remove some human biases, such as mood, fatigue and rapport, but they can introduce others if the scoring model learned from biased historical decisions or penalizes accents and speech patterns. Treat AI tools as another selection procedure: ask vendors for validation and adverse impact data, keep a human reviewing decisions, and check whether local rules such as NYC Local Law 144 require a bias audit and candidate notice. Rules vary by jurisdiction.
Frequently asked questions
What is interview bias?
It is a systematic error in how interviewers judge candidates, where factors unrelated to the job, such as first impressions or shared background, influence ratings and decisions.
What are the most common types of interview bias?
First-impression bias, the halo and horns effects, affinity bias, confirmation bias, contrast effect, central tendency, recency bias, nonverbal bias and stereotyping.
How do you reduce bias in interviews?
Use structured interviews with pre-written rating anchors, score each competency independently from notes, use diverse panels, limit unscored small talk and audit pass rates by group.
Is interviewer bias illegal?
Bias itself is a psychological tendency, but decisions influenced by race, sex, age, disability or other protected characteristics can violate laws such as Title VII, the ADEA and the ADA, whether or not the bias was intentional.
Does unconscious bias training remove interview bias?
On its own, the evidence that it changes hiring outcomes is weak. It works better paired with structural changes like standardized questions, scoring anchors and independent ratings.