Interview bias

How to avoid interview bias in your recruitment process

Spread the love

What’s the most pointless interview bias question you’ve been asked? For me, a few of the top contenders would be:

  • What football team do you support? (spoiler: I don’t follow it at all)
  • Have you been travelling?
  • Where do you stand on question marks?

Your traditional interview is prone to bias. At best, interviewers try their best to be objective whilst unknowingly being influenced by biases they didn’t even know they had. And at worst, interviewers will ask prying, personal questions to determine whether or not someone would be any fun down at the local pub on a Friday evening.

So, let’s look at how bias can impact on interviewing and what we can do to make interviews less biased and more data-driven…

Fact: Unconscious bias affects interviewing 

Indeed’s survey found that 28% of employers say ‘gut feeling’ is their main reason for hiring someone. The truth is: ‘gut feeling’ is essentially unconscious bias. Therefore, if you rely on ‘gut-instinct’ in interviews, then you’re probably being biased.

According to an study, about 5% of interview decisions were made within the first minute and nearly 30% within the first five, not long enough to make a fair, objective decision, I’m sure you’d agree. Our gut is often wrong since it relies on the intuitive, mental shortcut-based part of our brain as opposed to the slower, more conscious part.

These are just some of the unconscious biases which affect interviewing:

  • Stereotype Bias: When we assume someone has certain traits because they are a member of a certain group. 
  • Affinity Bias: We tend to gravitate towards those most similar to us.
  • Confirmation Bias: We search for information that confirms our preconceptions. 
  • Halo effect: Our judgment on one particular aspect of something can affect how we perceive other aspects. 
  • Groupthink: We may suppress our own objections in favour of group harmony.

Although some veteran interviewers might actually pride themselves on their magical talent-spotting gut, that feeling is most likely one of the biases above being triggered. The quick-fire associations and mental shortcuts our brains make mean that candidates are not judged fairly, and hirers miss the best talent. 

In fact, these biases are probably the reason why traditional, unstructured interviews have been proven to be one of the worst predictors of job performance – because decisions are not based on skill, but bias.

The only way to remove bias is by redesigning our environment 

The evidence shows that just making people aware of bias doesn’t change behaviour. A study of 829 companies over 31 years showed that diversity training had no positive effects in the average workplace. The reason training alone can’t de-bias interviewing (or hiring as a whole) is that it tries to change people. 

Since unconscious bias is part of being human, it would take years of constant training to make any tangible change – that’s if it’s even possible at all. Instead, we can simply design better choice architecture. Whilst we can’t change poople, we can change the environment in which they make choices. And so, by re-thinking how we interview, we can reduce bias. Once you remove bias, your process will become more objective (and therefore more predictive) as a result.

How to run more data-driven interviews

Conduct structured interviews

The first step towards fairer interviewing and recruitment process is asking all candidates the same questions in the same order. Most interviews tend to be go-with-the-flow affairs, which mean that not all candidates are given an equal chance to shine or even demonstrate the same skills.

If the interview is based around taking detours and delving into the candidate’s background, every interview will be completely different and therefore impossible to compare. It then just becomes a matter of who you got along with the best, and not who was actually best equipped to do the job.

Ask the right questions

Putting too much emphasis on education and experience could be a mistake. These are actually poor predictors of ability:

Image refering to the Study: The Validity and Utility of Selection methods in Personnel Psychology (by Schmidy & Hunter) image property of

Not only are education and experience bad at predicting how capable someone really is, but asking candidates about their background also allows for bias to enter the process and influence decisions. This doesn’t mean that education and experience are worthless – of course they’re not.

However, you should be looking to test for skills learned through experience and not experience itself – since not all experience is equally valuable. It may also be the case that someone’s skills gained through experience outside of your industry or at an unknown company is actually what makes them the best person for the job.

So how do you effectively test for skills?

According to the chart above, the answer is ‘work samples’. These are job-specific questions designed to simulate the job itself.

What could be more predictive than getting candidates to essentially perform parts of the job? To create your own work sample-style interview questions, start by shortlisting 6-8 essential skills needed for the role.

Then for each of these skills, think of a task or even a real-life project that would test this skill, should the candidate get the job. Ask for real situations in their past where they had to face a similar situation to what would happen at your company. Base the interview in competency based questions and examples.

If a candidate can do the job, then that’s all that should matter. Taking the emphasis off of education and experience isn’t lowering the bar – it’s simply removing triggers for bias.

Scoring criteria

All of your interview questions should have scoring criteria. This criteria doesn’t need to be too detailed, it can be as simple as a 1-5 star scale. Just note what points or condsiderations you think an average, good and great answer should include. Not only does scoring criteria mitigate against some bias, but it also means that once you’ve interviewed all candidates, you can build a leaderboard that will clearly show who the best candidate is – and you’ll have the data to prove you hired the best person.

When scoring answers against your criteria, try to judge each answer in isolation. Why? Because there are ordering biases at play that may result in candidates being scored higher or lower than they should be based on answers to previous questions.

For example, if a candidate answers question 3 perfectly, you’ll be likely to score question 4 more favourably than you would’ve otherwise. To give you an idea of what your criteria could look like, take a look this example:

Screenshot of an in-app for a role that hired for.

Crowd wisdom

Crowd wisdom is the general rule that the judgment of a group tends to be more accurate than that of an individual. By applying this theory to interviews, you can negate some individual biases by having multiple interviewers. Three interviewers (even if is not in the same interview) is the ideal number, since any more than that and you’ll start seeing diminishing returns.

This is also why you need scoring criteria – so that colleagues, who might not have been privy to the screening stage or previous interviews, can still join in and help score candidates.

It goes without saying, that interviewers shouldn’t discuss scores or their opinions of candidates until after they’ve finished scoring – the easiest way to go about this is to simply score questions in the interview itself, or at least make notes for scoring afterwards. 

Putting it all together – bias-free interviewing checklist

If you liked this blog post about “interview bias”, you can receive our regular updates and get a free job description template by clicking below Yes, I want to receive your regular updates and get a free job description template!

Leave a Reply

Your email address will not be published. Required fields are marked *