Anytime we interact with someone new, our brains make dozens of tiny snap judgments.
Often, we aren’t even aware of these impressions, but when they’re based on factors such as race, gender, age or sexual orientation, they’re considered a form of harmful stereotyping called unconscious bias.
Unconscious bias is a more subtle form of discrimination that often flies under the radar, but the recent groundswell of support for the racial-justice movement has shone a light on unconscious bias in the hiring process. Women and minorities are still wildly underrepresented in leadership and technology roles, and people with disabilities are not even being included in most companies’ diversity reporting.
New tech platforms have sprung up promising to remove bias using AI or by making the entire hiring process anonymous.
But the unfortunate truth is that there’s no single app that can eliminate bias. The only way to be more inclusive is for leaders to reevaluate their current hiring practices and work to eliminate bias one step at a time.
Here are a few ways to work toward removing bias from your hiring process.
1. Expand talent-sourcing channels
Most people don’t even consider the fact that where they’re sourcing candidates could be introducing bias to their hiring process. But if you always attend the same conferences, job fairs or alumni events, you could be skewing your pool of potential candidates. Certain universities or geographic locations can favor people of a specific socioeconomic background or race.
To combat this issue, try sourcing candidates from a broader range of sources, such as social media sites, job boards or career fairs in another part of the city. You might be impressed by the number of qualified candidates you find simply by casting a wider net.
2. Reevaluate how you screen applicants
Eliminating bias starts long before candidates ever sit down for an interview. Your job posting and the way you vet applicants could be preventing some of your most qualified candidates from ever getting a foot in the door.
Studies have shown that women will only apply for a job when they meet 100 percent of the listed criteria, while men will apply even if they only meet 60 percent of the requirements.
Listing certain skills or job experience as “required” (even if they’re just a bonus) could deter many great candidates from even submitting an application. Make sure the posting accurately reflects what the job requires, and stay away from highly gender-charged words.
Once it’s time to review applications, resist the urge to shuffle all the candidates from prestigious schools to the top of the stack.
Top universities are rife with their own biases and generally favor kids with privilege. Cherry-picking candidates from these selective schools just strengthens these imbalances and reduces the chances of a non-white hire by 23 percent.
3. Don’t rely on shortcuts
To get candidates through the talent pipeline more quickly, many companies have begun using AI software.
Some platforms designed for screening software engineers claim to be able to detect cheating by recording and analyzing candidates’ faces.
HireVue, a video interview platform, has recently come under scrutiny for using facial movements, voice and word choice to gauge a person’s mental agility and emotional skills.
While using AI software to analyze candidates’ résumés or biometric data might sound appealing, it can actually do more harm than good. For one thing, these platforms are not without bias and may even thwart efforts toward greater inclusivity.
AI tools are only as good as the data that trained them, and the tweets generated by Microsoft’s Tay and OpenAI’s GPT-3 show that any AI created by humans will have humans’ flaws and biases.
Software designed to score candidates based on their biometric data will only reinforce biases toward candidates who are already likely to be hired.
4. Use assessments that test candidates’ job skills
Back in the early 2000s, Google became famous for its insanely difficult interview riddles. Google eventually phased out these brainteasers because they weren’t an effective means of identifying great candidates.
When it comes to finding the right person for the job without inserting unconscious bias, it’s best to give candidates objective assessments that reflect real-world job duties.
For a digital marketer, this could mean asking him or her to create a high-converting Facebook ad campaign; for a software engineer, it could be a take-home project.
Coderbyte is one platform that aims to help technical recruiters evaluate candidates more thoughtfully. It offers automated code assessments for screening, interviewing and final take-home projects without AI overreach.
For example, the platform actually encourages candidates to conduct Google searches from within the platform itself rather than deploy face recording or other potentially biased methods to detect cheating.
This more closely resembles how developers actually write code, thereby allowing hiring managers to understand how candidates think and troubleshoot as they would in the day-to-day of their job.
5. Train your team to conduct unbiased interviews
While many hiring managers prefer to “wing it” in interviews, improvisation can actually introduce bias. Research shows that people naturally set a more casual tone for people they feel are part of their in-group.
This can lead to an interviewer asking people like him or her easier questions or making those candidates more comfortable with small talk.
To avoid this issue, draw up a list of relevant questions ahead of time and screen those for potential bias.
Then instruct the employees conducting the interviews to ask every candidate the same series of questions.
No one is immune to unconscious bias. We all make snap judgments about other people based on factors that have nothing to do with their skills and personality.
The problem comes when bias creeps into the hiring process. It can shut out some of your very best candidates — the people whom you want on your team. There are no shortcuts to removing bias, but we can all do more to unearth and eliminate it.