4 Tips to Remove Unconscious Bias From the Hiring Process

That's not a valid work email account. Please enter your work email (e.g. you@yourcompany.com)
Please enter your work email
(e.g. you@yourcompany.com)

StereotypeUnconscious bias and negative stereotyping can do serious damage to the equality and fairness of the modern hiring process. For example, a recent study found that  men are twice as likely to be hired for a mathematical task than women, even if the men and women appear equally qualified. This is likely only the tip of the iceberg, and unconscious bias is thought to negatively impact hiring right across the board.

In this article, rather than focus on the mechanics and negative impact of unconscious bias, I want to outline the steps that can be taken to remove unconscious bias from the hiring process.

1. Removing Names From Resumes

An editorial published last September in Scientific American  argues that one way to “fight race and gender bias in science” could be to remove identifying information, such as names, gender, etc., from resumes. Candidates can’t be unfairly discriminated against on the grounds of gender and race if they are only identified by a number.

I assume that most ATSs could be easily programmed to obscure names at key points. If not, removing names could be done manually. But, either way, the question is: would it work? The Scientific American editorial cites two studies in which data such as name, age, and gender were removed from job applications. These studies showed that women and minorities were more likely to make it to the interview stage under these more anonymous hiring circumstances, suggesting that, yes, removing identifying information could make the hiring process fairer and more equal.

2. Letting Computers Do the Selection

One advantage of computers is that they don’t have a subconscious and simply don’t carry all that unconscious baggage around with them like mere mortals do. This arguably makes them bulletproof against claims of stereotyping and bias.

Of course, we don’t have AI recruiters sitting in offices who can do the hiring, but we can use automated psychometric tests to assess candidates’ knowledge and skills. We can then develop a fair selection algorithm that uses the results of these tests to create candidate short lists — and possibly even appoints candidates to positions for which they are qualified. This already happens at Xerox, which uses a computer algorithm and computerized psychometric assessments to hire candidates. The company cut retention rates by a fifth hiring this way. It would be interesting to see whether the Xerox process led to a reduction in unfair discrimination in hiring as well.

3. Implicit Bias Testing

It would be depressing to think that humans cannot be trusted at all to hire fairly and that we should just hand hiring over to the robots. Luckily, it’s not quite that simple. In fact, it seems that humans can have varying levels of negative bias, ranging from negligible to outright disdain.

The good news is that we humans can use self-assessments like the Implicit Association Test from Harvard  to discover our own levels of bias. Moreover, once we are aware of our biases, we can take steps to correct them. If one has been conditioned to be unfairly negatively biased, one can also condition themselves to become more positively biased. It is perfectly possible for organizations to put hiring teams and hiring managers through diversity awareness programs in order to reduce unconscious bias significantly.

4. Blind Interviewing 

Research shows that  blind auditions have helped increase the number of female musicians in orchestras by removing both explicit and implicit gender discrimination. It’s one thing to implement blind auditions in orchestras, and it’s another thing to do it when hiring candidates, but surely this option can be explored.

That being said, problems do exist with blind interviewing in the corporate sphere, as interviewers can still determine peoples’ genders via the sound of their voices, and some interviewers hold unconscious biases about certain accents.

By Kazim Ladimeji