Diversity at Work: A Guide to Reducing Bias in Hiring

What is Unconscious Bias? 

Biases are useful in many ways, as our brains must process millions of pieces of information at any one time, most of this processing happens unconsciously. In other words, we’ve created mental shortcuts for ourselves.   

So, when we make a quick assessment or judgment about other people – which we do every time we see or interact with people – most of it happens unconsciously based on our background, cultural environment, and personal experiences. It’s fast, automatic and, as the name suggests, unconscious. 


Everyone With a Brain Is Biased 

It’s easy to feel offended when someone tells us we’re biased. Most of us like to think we’re open-minded, rational, enlightened people. Bias has a stigma attached.  

The fact is that regardless of our very best efforts, scientists have shown that every single brain has over 100 biases built in. They’re built into our brain structure and are a critical part of how we process information. In short: if you have a brain, you have biases. 

Even if you think you’re not biased, there’s a reason for that. It’s called Blind Spot Bias, also known as “naïve realism,” and is one of the most challenging biases to overcome. It leads us to believe that our own perception of reality is objective, accurate, and not influenced by other biases.  



How Do Biases Impact Our Behavior? 

Some claim we have over 150 different biases embedded in our brains; others place the number around the 100 mark.  

On a practical level, biases can come into play when we’re hiring new people, conducting performance reviews, bringing teams together, and offering development opportunities. They can also impact team performance, decision-making effectiveness, organizational culture and, ultimately, the success of the organization. 

Let’s look at few examples. 

Example 1: Gender Biases 

A study from Yale University asked male and female science researchers to rate two candidates with the same qualifications for a lab manager position. All of the researchers rated the male candidate more favorably than the female candidate. They were also willing to pay him a higher salary. 

Example 2: Names/Ethnicity 

Numerous studies have demonstrated that resumes with ethnic sounding names are less likely to be offered an interview. For example, 

  • One study found that Black-sounding names received 50% fewer callbacks compared to identical resumes with white-sounding names. 

  • “Whitened” resumes that remove details indicating ethnic group membership were more likely to receive callbacks. Black applicants who whitened their resumes were 85% more likely to get a callback, while Asian applicants who whitened their resumes were 58% more likely to get a callback. 

During a recent study conducted by the Australian National University, over 4,000 resumes were sent out to different organizations. Each was identical, apart from the candidate’s name. They found that to get as many interviews as an applicant with an Anglo sounding name: 

  • An indigenous person had to submit 35% more applications; 

  • A Chinese person had to submit 68% more applications; 

  • A Middle Eastern person had to submit 64% more applications. 

Example 3: Voices/Accents 

Associate professor of linguistics at Stanford University, Meghan Sumner, found that where we’re from and how we feel about particular accents dictate how we listen to others. Most of us prefer a particular type of voice in terms of gender, accent, and intonation.  

And, despite MRI data analyses showing that bilingual people have better cognitive processes including attention and inhibitory control, memory and cognitive flexibility, than monolingual people, many people with English (or any other language native to the country they now live in) as a second language find themselves discriminated against because of their accents. 


21 Harmful People Biases That Can Influence How We Hire 

  1. Blind Spot Bias/Negative Realism - This is one of the most difficult biases to overcome: when we think our perception of reality is objective and accurate and not influenced by biases.  

  1. Stereotypes - One of the more well-known types of bias, this is when we form over-simplified opinions about a person or group that doesn’t take individual differences into account, often based on gender, ethnicity, age, or weight. 

  1. Superiority Bias - When we don’t know much about a particular topic, we’re quite likely to believe that our judgment is less biased and more acute than other people’s. This bias is also known as illusory superiority or the Dunning-Kruger Effect. 

  1. Contrast Bias - This happens when we form an opinion based on the standard of what came previously.  

  1. Confirmation Bias - We interpret new information we receive in a way that confirms our original beliefs and tend to ignore information that contradicts them.  

  1. Anchoring Bias - The classic example of anchoring bias is price negotiations: the first price quoted or offered will set the range for (or anchors) the rest of the negotiations. 

  1. Framing - We can draw different conclusions about the same piece of information depending on how, or by whom, it’s presented. The media, advertisers, and politicians use this tactic frequently to persuade people to change their minds about an issue. 

  1. Halo/Horns Effect - This is a sub-type of the confirmation bias and refers to having a positive (or negative) first impression of someone or something, which then leads us to interpret subsequent events positively (or negatively). 

  1. In-Group vs. Out-Group Bias - We tend to prefer people who belong to the same group as us, with a shared interest, identity, or characteristic.  

  1. Group Attribution Error - We believe that the characteristics of one person represent the whole group. So, for example, if the only female engineer we’ve ever worked with is very efficient and well-organized, we might think that all female engineers will have the same characteristics. 

  1. Availability Heuristic - Often, we believe that things we can recall easily are more frequent or important than they really are.  

  1. Gambers’ Fallacy - This occurs when we think that future probabilities are altered by past events – such as believing we’re more likely to get heads on a coin toss if the past five events have landed on tails.  

  1. Recency Bias - We usually assign more weight or value to recent events or information rather than earlier events. 

  1. Actor Observer Bias - We often think that the way other people behave is based on their personality, rather than situational influences.  

  1. Pro-Innovation Bias - Sometimes, if we strongly believe in a particular innovation, we believe that everyone should adopt it as it is, without taking the time to analyze any possible shortfalls or limitations. 

  1. Affect Heuristic -  Affect heuristic makes it so that if you are having a bad day, you may also have a “bad feeling” about a candidate, even if it has nothing to do with their qualifications for the job. 

  1. Bandwagon Effect - We’re often influenced by the choice of the majority, even if it goes against our own judgment. As more people accept a belief or point of view, the more likely we are to accept it as well.  

  1. Commitment Bias - Once we’ve incurred expense or invested in something, we’re less inclined to let it go, despite evidence to the contrary.  

  1. Outcome Bias - We can often judge a decision based on the eventual outcome, rather than the process we used to reach that decision at the time (particularly if the process was sub-par).  

  1. Social Comparison Bias - We might be inclined to dislike or want to compete with someone we perceive as mentally or physically better than us. So, when hiring, we could favor people who we think are not as good as us, so we don’t feel threatened or overshadowed by them. 

  1. Egocentric Bias - This is one to look out for in candidates: the tendency for people to claim that they were solely or primarily responsible for an outcome that was actually the result of a joint activity. 


How to Mitigate Unconscious Bias

Adopt a highly structured recruitment process 

A best practice recruitment methodology looks something like this: structured, consistent, objective, and designed to remove bias and enhance diversity at every step of the process. 


STEP 1 | Write an Inclusive Job Description 

While the job description might seem like the most innocuous part of the hiring process, it can still leave room for a significant amount of bias. The job description is so critical because it represents the first touchpoint that a candidate has with an organization.  


STEP 2 | Source From Multiple Channels 

The more candidates you can get your job in front of, the better your chances at attracting top performers.  

Beyond just posting a job to different forums, organizations can also take a more proactive approach to attracting a diverse candidate pool. Recruiters may specifically reach out to certain sources, such as: 

• Professional organizations dedicated to serving diverse groups 

• Traditionally black colleges and universities 

• Women’s colleges and universities 

• Alumni associations 

• Networking groups 

This can increase the number of eyes that see the job description and can take a more targeted approach to encourage applicants from specific backgrounds to apply. 


STEP 3 | Screen Candidates Consistently 

The whole purpose of a recruitment process is to identify the most suitable candidates for a role and have them fill out an application.  This is easy if you use an Applicant Tracking System (ATS) or a job board that allows you to ask questions of your candidates. However, if you don’t have access to either of these, the next best thing is a simple checklist, which could include: 

  • A list of the absolute essentials for the position, based on the job description  

  • A list of the “nice to have” criteria, based on the job description with a checkbox or rating scale for each one 

  • A rating (e.g. 1-3) for how polished the application is (especially if the role requires a high level of literacy or written communication skills).  

At the end of this process, you should have a list of checks in boxes and ratings. You can assign a “1” for each check, and add this and your ratings together. You can then sort your applications into three groups: Yes, No, and Maybe. 


STEP 4 | Assess Candidates Objectively & Scientifically 

A meta-analysis conducted by Gallup found that when companies selected the top 20% of candidates using a scientific assessment, they were likely to experience 41% less absenteeism, 70% fewer safety incidents, 59% less turnover, and 21% higher profitability. 

What is surprising is that some organizations still aren’t convinced of the effectiveness of pre-employment assessments, despite all of the evidence that confirms they’re the most efficient and effective way to identify top performers. 

By incorporating scientifically valid, predictive assessments, you can highlight diverse candidates who have not yet had the opportunity to perform in a certain field or industry. The result is increased access and a more diverse set of applicants from which to choose. 


STEP 5 | Conduct Structured Interviews 

Unfortunately, many hiring managers prefer unstructured interviews – an informal, “getting to know you” type of process that allows them to explore different facets of each candidate – over a more structured, scientific approach.  

It’s unfortunate because unstructured interviews have been repeatedly shown to be one of the very worst methods of predicting performance at work. 

The best – and most scientifically reliable – strategy at this stage is a structured, standardized interview process with identical questions, used consistently across every candidate. 

Bohnet suggests several other strategies to increase objectivity for interviews: 

  • Questions should be asked in the same order for all candidates. 

  • Each interviewer should score each question immediately after the candidate answers it.  

  • We should compare responses horizontally when reviewing them. So, for example, if we interview ten candidates, we should review each candidate’s response to Question 1 at the same time, then Question 2 and so on. 


STEP 6 | Conduct Reference Checks 

Like interviews, reference checks can be inadvertently influenced by biases. The best strategy is like conducting structured interviews: each referee should be asked the same set of questions; their responses should be recorded and scored at the time; when reviewing we should ensure we have sufficient evidence to make a judgment. 


In Conclusion 

“Blind” Recruitment 

Recently, the Australian Bureau of Statistics (ABS) wanted to increase the numbers of females in senior roles, so they removed names, genders, and other identifiers from applications. In addition, they emphasized flexible hours and working from home options as well as providing training for their interview panels and management on unconscious bias. As a result, 15 of 19 senior hires were female, whereas previously only 21% of their senior roles were filled by women. 

However, the effectiveness of blind recruitment warrants further research (particularly in psychological literature) to determine the overall effectiveness of other initiatives. It may be helpful, and a good step towards decreasing the impact of unconscious bias, but it’s not a silver bullet by any means.