Best Practices to Reduce Hiring Bias

It’s time we address the elephant in the room — your company needs help increasing its diversity. It’s no secret that more diverse companies have a competitive advantage, and not just in terms of diversity of thought. More diverse companies have cost advantages, improved resource acquisition, greater marketing ability, system flexibility, and enhanced creativity and better problem solving.

What company wouldn’t wanna get in on this!?

While managing diversity can be a confusing and complicated process, we have identified one major barrier to increasing your company’s diversity. Hiring bias.

What is hiring bias?

One of the most common biases we see from hiring managers is “affinity bias,” which is a favorable opinion toward candidates who look, act, or operate in a similar fashion to that of the hiring manager. Unfortunately, the human mind is pre-programmed to unconsciously side with candidates who remind us of ourselves. This in turn leads to confirmation bias, supporting our existing beliefs about the candidates we interview. Similarly, the halo effect will lead us to get a positive impression of a person based on a single characteristic that we find attractive like looks, voice, or a shared interest.

Undoing these unconscious behaviors is not an easy task and requires a conscious effort to overcome the personal bias. Accepting this unconscious bias is the first step toward making sure that equal opportunities are presented to every candidate. As the hiring manager, you must ask yourself questions about the possibility of bias and where it might show up. It is also essential to remove any groupthink patterns which could influence your thoughts about certain candidates and how well they might fit into the company culture.

Steps to Combat Hiring Bias

Luckily for you, we did the research for you! Here are some easy, actionable steps your hiring managers can take to reduce hiring bias.

1. Insist on a diverse pool
2. Use software that blinds demographic characteristics.
3. Narrow down your list of candidates with phone interviews first, but beware of voice bias.
4. Utilize highly structured interviews
5. Randomize and split evaluations
6. Use specific, defined metrics when rating a candidate’s skill
7. Pre-commit to objective criteria

Below we’ll dive into each of these steps.

In hiring, leaders should insist on a diverse pool

Before you even get into the interview process, you need to work on increasing the diversity in your pool of candidates. Start by reviewing your job postings. Is the language inclusive? Do you find that a lot of the required soft skills are gendered? Is the listing full of jargon?

Next, really think about whether or not the job requires a degree. If your open position is an entry level customer service position, perhaps somebody who has worked in the service industry could have the same skills needed for this job and a degree may not be necessary. If your job does require a degree, consider using software to blind the candidate’s university — not everybody has the privilege to access Ivy League schools, and favoring these kinds of institutions can lead to bias.

Traditional recruiting channels like LinkedIn and Indeed can be an awesome resource for recruiters, and you should definitely keep using them. However, if you’re trying to target less represented demographic groups, go beyond these channels. Try using one of the following:

• Diversity.com
• Black Career Network
• Hispanic/Latino Professionals Association
• Workplacediversity.com

These channels will specifically target diverse groups of qualified candidates and are definitely worth looking into.

Use software that blinds demographic characteristics.

Browsing a candidate’s LinkedIn or social media profile’s is a great way to find out more about their skills, educational background, job history and qualifications. However, we risk bias when we view their demographic characteristics.

So this one is easy.

There is free software that can blind a candidate’s demographic characteristics.

Try Unbiasify Chrome Extension! It allows you to hide names and photos on LinkedIn, Facebook or Twitter to reduce the unconscious bias based on demographic characteristics, like name or gender.

Anti-bias is another free extension that blinds a LinkedIn profile’s characteristics.

Using these software will help you peruse your candidates profiles more objectively.

Narrow down your list of candidates with phone interviews first, but beware of voice bias.

Remember that pesky halo effect and affinity bias we talked about? Using phone interviews to narrow down your candidates can help combat bias based on appearance, because obviously you can’t see what they look like. However, research shows that we can develop bias based on voice, tone and even accents. So if you are using phone interviews, keep these things in mind to combat voice bias:

1. Critique the content of the conversation, not the tone of voice.
2. Critique preparation. Do they seem engaged? Did they ask interesting follow-up questions?
3. Be aware of the risk of bias and try not to attribute a voice to intellect, educational background or any demographic characteristics.

Utilizing these methods should reduce appearance based bias, while being aware of voice bias should help limit it as well.

Utilize highly structured interviews

Do you ever shoot the breeze with your candidate and realize that you both are superfans of the tv show Game of Thrones? And now you’ve bonded over how terrible that final season was, (I’m still not over it.) Sorry to break it to you, but that warm and fuzzy feeling towards your candidate is your bias. You’re not supposed to evaluate a candidate on whether or not you like them, but on whether or not they are the best fit for the position or the most qualified candidate.

It’s ok. You’re only human. A highly structured interview can help.

In a structured interview, the interview asks a set of questions in a standardized order then grades the candidate against a predetermined scoring system. The scoring system should include defined hard and soft skills. A soft skill might be “excellent communicator”, which should be defined in order to avoid bias, like accent bias for example. This also helps to eliminate similarity bias and keeps interviews consistent.

Question types can include:

• Job-specific structured interview questions (i.e. “What is your preferred programming language and why?”)
• Behavioral structured interview questions (i.e. “What has your biggest professional challenge been?”)
• Situational structured interview question (i.e. “What would you do if you were assigned to work with a difficult client?”)

This method ensures that each candidate receives roughly the same amount of allotted time for their interview and ensures that the conversation does not deviate away from interview related questions. Sorry GOT.

Randomize and split evaluations

While a full look at an applicant’s profile is essential for a good evaluation, you may let your perception of the whole application get weighted by, or anchored on, the information presented in the start, a phenomenon known as anchoring bias (for example, knowing a candidate went to Stanford could make you overrate lackluster work experience). Therefore, before the crucial big-picture look it helps to separate qualifications and analyze them in random order before our view of them becomes biased by the information that surrounds it.

Use specific, defined metrics when rating a candidate’s skill

In order to focus on relevant attributes and not get distracted by irrelevant ones, we should create well-defined metrics for how these attributes should be evaluated. Make a list of desired characteristics in each area and rate how the application fulfills these requirements. If needed, include metrics that require you to look at their entire application to rate, but make sure to evaluate these last.

Pre-commit to objective criteria

Not every metric evaluated will be objectively measurable, but wherever possible, use objective criteria so that evaluations will not be dependent on factors such as mood at time of review or word choice of the applicant, to name a couple.

These measures should be applied, as permitted, to every step of the application process. To help inspire you in how to implement them and make the concepts more tangible, we made a step-by-step example of how you may incorporate them to resume review for a software engineer position:

Step 1: Decide on the Parts to Evaluate

This should be information included on most resumes in your field. For our fictitious engineer position we will use:

Education Work experience Skills/Proficiencies

Step 2: Decide on Evaluation Criteria

Education:

• Studied AI/Machine learning
• Studied web design

Work experience:

• Worked on big data for 3 or more years
• Has seen work on data models from beginning to end

Skills/Proficiencies:

• Can write in software on our data stack (JS, React, Python)
• Has experience in a variety of AWS services
• Can write in software on our data stack (JS, React, Python)
• Has experience in a variety of AWS services

Your evaluation scale can be as simple or complex as needed, so long as you make it clear ahead of time. Our scale will go from 1 to 5, where 5 is ideal, 3 is solid, and 1 is total mismatch.

Step 3: Obscure Candidate Names

Have a helper assign codes to each candidate which will only be shared after your review is complete. As part of this step, they should remove personally identifying information that might bias your decision-making such as age, gender and family status.

Step 4: Split Candidates’ Information

Have a helper assign codes to each individual part so they can be separated to obscure which section belongs to which candidate.

Hidden: 352 191 775 -> X

Step 5: Evaluate by Parts

Evaluate each application part according to the metrics defined.

Step 6: Put Parts Back Together

Once we have evaluated each individual part without bias from the others, we put them back together using our reference. At this point we can give the application full-picture ratings.

Overall rating: 4

Step 7: Decide on Who to Contact

After reviewing all applications, we decide on who to contact for the next step of the application process, and only then share the selected candidate’s name or other identifying information.

Final Thoughts

Fixing your hiring bias is an actionable way to start managing diversity better at your company. Utilize some of these tactics and you’ll be targeting more qualified, diverse candidates in no time!

Check out our references and learn more about hiring bias here:

Bounds, Darren. “3 Rules for Using Inclusive Language in Job Ads — Breezy HR.” RSS, 2018, breezy.hr/blog/3-simple-rules-for-using-inclusive-language-in-your-job-ads.

Huang, Laura, Frideger, Marcia, and Pearce, Jone L. “Political Skill: Explaining the Effects of Nonnative Accent on Managerial Hiring and Entrepreneurial Investment Decisions.” Journal of Applied Psychology 98.6 (2013): 1005–017. Web.

Miller, Gail. “Judging the Voice: The Reality of Phone Interview Bias.” ERE, 23 July 2015, www.ere.net/judging-the-voice-the-reality-of-phone-interview-bias/.

Morgan, Whitney Botsford, Walker, Sarah Singletary, Hebl, Michelle Mikki R, and King, Eden B. “A Field Experiment: Reducing Interpersonal Discrimination toward Pregnant Job Applicants.” Journal of Applied Psychology 98.5 (2013): 799–809. Web.

Stewart Black, J., et al. “5.3 Diversity and Its Impact on Companies — Organizational Behavior.” OpenStax, 5 July 2019, openstax.org/books/organizational-behavior/pages/5–3-diversity-and-its-impact-on-companies.

“Structured Interviews: What Are They and Structured Interview Questions.” Indeed Career Guide, 2020, www.indeed.com/career-advice/interviewing/structured-interviews.