An interview is a judgement made in under an hour, on limited evidence, by someone who usually has their own work waiting. Under those conditions every human mind takes shortcuts. The shortcuts are not random. They follow well-documented patterns that psychologists have studied for decades, which is good news: predictable errors can be designed against.

Below are the biases that most often distort hiring decisions, each with what it looks like in an Indian hiring context and what actually reduces it.

First impressions and the rest of the hour

First-impression bias is forming a view in the opening minutes, from handshake, accent, clothing or small talk, and then spending the rest of the interview unconsciously looking for confirmation. That second step is confirmation bias. An interviewer who decided early that a candidate seemed “not confident” will read a pause before answering as hesitation; the same pause from a candidate they liked is “thoughtful”.

A related effect is the halo or horns effect, where one strong trait colours everything. A candidate from a well-known IIT gets the benefit of the doubt on a weak system design answer. A candidate who stumbles on the first technical question is judged harshly on everything after. In both cases one data point is doing the work of many.

What helps: deciding before the interview which competencies will be assessed and with which questions, and scoring each answer separately as it is given, before forming an overall view. Writing down the evidence for each score (“explained cache invalidation trade-offs with a concrete example”) rather than impressions (“seemed sharp”) makes it harder for one moment to dominate.

Liking people who resemble us

Similarity or affinity bias is warmth towards candidates who share our background: same college, same hometown, same language, same previous employer, same hobbies. It is often mislabelled as “culture fit”. The candidate who also grew up in Coimbatore and supports the same IPL team creates easy rapport, and rapport feels like evidence of suitability. It is not.

Over many hires, similarity bias makes teams more homogeneous, which narrows the range of perspectives and, in India, can quietly reproduce regional, caste, gender or college-tier patterns without anyone intending it. A practical fix is to define culture in terms of observable behaviours the organisation values, such as how someone handles disagreement or owns mistakes, and assess those with specific questions, rather than relying on a “would I enjoy working with them” feeling.

Comparing against the last candidate

Contrast effect is judging a candidate relative to whoever came just before, instead of against the standard for the role. On a campus drive where an interviewer sees fifteen students in a day, an average candidate following two weak ones looks excellent, and the same candidate after an outstanding one looks mediocre. Order effects also include fatigue: decisions late in a long day tend to be less careful.

What helps: rating anchors that describe what a 2, 3 and 4 look like for each competency, so candidates are compared with a fixed description, not with each other. Limiting the number of interviews per panellist per day and building in breaks also reduces fatigue effects.

Stereotypes and assumptions

Stereotyping applies assumptions about a group to an individual: that a woman with young children will not handle travel, that a candidate over 45 will not learn new tools, that someone from a tier-3 college cannot handle complex work. These assumptions often surface as illegitimate questions about marriage, pregnancy plans or family responsibilities. Beyond being unfair, they create legal exposure; Indian law prohibits discrimination in recruitment on several grounds, including sex under the Code on Wages and disability under the Rights of Persons with Disabilities Act.

What helps: asking every candidate the same job-related questions, and if a genuine requirement exists (such as frequent travel or night shifts), stating it and asking every candidate whether they can meet it, rather than guessing for some.

Group effects on the debrief

Bias does not stop when the interview ends. In the debrief, the most senior person often speaks first and others adjust their views to match. This is anchoring combined with conformity. A panellist who had doubts but hears the director say “strong hire” may stay silent.

What helps: every panellist submits written scores and evidence in the ATS before the debrief, and cannot see others’ feedback until theirs is in. In the discussion, the most junior panellist speaks first. Disagreements are resolved by looking at evidence for specific competencies, not by voting on overall impressions.

Why training alone disappoints

Many organisations respond to bias with a one-time awareness workshop. Awareness is useful, but research on unconscious bias training generally finds limited lasting effect on actual decisions when it is not backed by process changes. Knowing that the contrast effect exists does not stop it from operating at the twelfth interview of the day. The interventions with better track records change the structure of the decision:

  • Structured interviews with the same questions and anchored rating scales.
  • Independent scoring before group discussion.
  • Diverse panels, so no single perspective decides.
  • Work samples and job-relevant tests that produce comparable evidence.
  • Reviewing outcomes: pass rates by gender, college tier or other groups at each stage, which can reveal a biased step that no individual noticed.

The goal is not interviewers with no biases; that is not possible. It is a process where any one person’s bias has less room to decide the outcome.

Written by the RP Strativa Team. Scenarios described are illustrative.

Leave a Reply

Trending