How to Reduce Unconscious Bias in Recruiting Interviews

Unconscious bias training is the most common organisational response to bias in hiring — and also one of the least effective. Research on the impact of standalone bias training on hiring outcomes is mixed at best. The reason is that awareness of bias does not reliably change behaviour in the moment, especially in low-structure, high-cognitive-load situations like an interview. Structural interventions work better than awareness interventions.

Why awareness alone does not work

Unconscious biases operate precisely because they are unconscious. A person who has just completed a bias awareness session may be more vigilant in low-pressure situations but will revert to automatic processing under time pressure, emotional engagement, or cognitive load — which are all present in most interview situations. Training that increases awareness without changing the structure of how interviews are conducted produces limited durable change.

The structural interventions that do work

Structured interviews: The evidence for structured interviews as a bias-reduction tool is strong. Standardised questions, consistent scoring guides, and explicit criteria for evaluation reduce the influence of irrelevant factors (perceived cultural fit, communication style, physical appearance) by anchoring evaluation to job-relevant competencies. See our research summary on structured vs. unstructured interviews.

Diverse interview panels: Single-interviewer decisions are more susceptible to individual bias. Multi-interviewer panels with diverse perspectives, combined with structured calibration debriefs, catch more potential bias and produce more defensible decisions.

Blind skills assessments: For roles where specific skills can be assessed objectively (technical roles, writing-intensive roles, analytical roles), blind skills assessments conducted before the interview substantially reduce the influence of first-impression bias. The candidate who performs in the top quartile on a work sample test deserves a fair hearing regardless of how they interview.

The debrief: where bias most often enters

Perhaps the most underappreciated bias risk is the post-interview debrief. In unstructured debriefs, the first opinion stated tends to anchor the discussion — and the first opinion is often formed by the most senior person or the most confident speaker, not necessarily the most relevant evaluator. This is called the anchoring effect, and it can override independently formed assessments of multiple interviewers.

The fix is a structured debrief: each interviewer submits an independent written assessment before the group conversation, using a shared scoring guide. The group then discusses the assessments, with explicit focus on evidence rather than overall impressions. This process is more time-consuming but substantially more reliable.

The role of real-time support in interviews

AI coaching tools can help interviewers stay on-script in the moment — flagging when questions deviate from the structured guide, surfacing the next competency area when one has been covered, and noting when a follow-up probe is likely to yield more evaluable information. This kind of in-the-moment support is more effective than pre-interview training because it operates at the point of decision rather than before it. See how Koachly supports structured interviewing.