Human Resources > Selection
> Interviewer Bias
Rapid convergence |
Limited questioning | Limited thinking |
Interviewer personality | See also
Interviewing is a skilful and important activity that, if done wrong, can
result in rejection of good candidates or, worse, appointment of people who will
prove problematic in practice. Here are some of the ways that interviewers may
show bias during interviews.
It can be very helpful for interviewers to review possible biases during
their preparation for the interview. This may even include some simulations with
feedback from observers. Then in the interview, they should remain self-aware,
including noting how the interviewee responds to their questions.
There is a common tendency to leap to conclusions as early as possible, and
thereafter only pay attention to evidence that supports those conclusions. This
- Preconceptions: Building an image of the candidate before the
interview, for example based on aspects of their resume (eg. education,
experience, personal statement).
- First impression: Basing all subsequent perceptions on an
impression of the person gained in the first few minutes (or even seconds!)
of the interview.
- Stereotyping: Assigning candidates to classic or personal
stereotypes based on limited attributes. Classic stereotyping may include
ethnicity, religion, gender, ability, etc. Personal stereotypes may be based
on such such as hair length, clothing, musical preferences, etc.
- Minimum criteria: Just looking for minimum attributes that
suggest the person can do the job. This is common when you desperately need
someone or where there are very few candidates.
The interviewer may have a preference for certain
questions, and in
doing so may tilt the interview or otherwise gain a distorted view of the
- Unbalanced questioning: Asking too many
questions, or otherwise not using an appropriately balanced set of
questioning styles. For example there should be appropriate
without over-doing this.
- Question selectivity: Using only a subset of possible questions
based on convergence, personal viewpoints, or other factors.
- Question variability: Asking mostly different questions to each
candidate without reason, leading to a difficulty in comparing candidates.
- Leading questions: Asking questions that lead to limited answers
that may be sought (perhaps unconsciously).
- Checklist blinkers: Seeking to gain answers based solely on a
fixed checklist and not considering wider personal characteristics,
experience, potential, etc.
The interviewer may also be biased in the way that they think in general and
how their thoughts are biased by these modes of thought.
- Halo effect: Seeing the person positively based on one or a few
preferred attributes, for example concluding a person who went to the
'right' university is particularly suitable.
- Horn effect: The reverse of the halo effect, judging a person
wholly negative based on a single or few factors.
- Gut-feel: Judging the person based more on what you feel about
them than a more thoughtful and balanced reasoning.
- Contrast effect: Comparing the candidate with other (often
recent) candidates rather than a more stable standard.
Aspects of the interviewer's
lead them to be biased in a range of ways.
- Leniency bias: Being kind, forgiving the person and overlooking
- Strictness bias: Being overly harsh and unforgiving.
- Central tendency: Seeing everyone as average and not
distinguishing better from less suitable people.
- Similarity bias: Giving greater importance to aspects where the
candidate is similar to the interviewer in some way, such as having worked
at the same company.
- Attractiveness bias: Assessing the person more on how they look
(most commonly done by men when faced with an attractive women).
- Loquaciousness: Talking too much rather than listening and
- Rescuing: Trying to help the interviewee get to the right
- Cynicism: Not believing anything, easily judging the interviewee as
- Gullibility: Easily believing what you are told without seeking
evidence. Judging the candidate on how slickly they present themselves
rather than the evidence.