If You Haven't Removed Bias from Performance Reviews, Then You're Not an Inclusive Leader. Take These Steps to Protect Your Marginalized Employees. Performance reviews are critical to pay, promotion and assignment opportunities — but they often hurt employees in marginalized groups, thanks to biases. Here are some to watch out for and how to remove them from your reviews.

Performance reviews are riddled with bias. From assessing folks based on potential vs. actual performance to biased language in annual review write-ups, women, people of color and other underrepresented folks often have fewer chances to succeed because of systemic bias.

Performance reviews are critical to pay, promotion and assignment opportunities. It's often during these discussions that compensation and career decisions are made that affect workplace representation and inclusion long-term.

Inclusive leaders keep their radars up for these biases, interrupt them when they see them and work to eliminate biases in performance reviews systematically.

Related: 7 Ways to Remove Biases From Your Decision-Making Process

Biases to watch out for

In the book Bias Interrupters, Joan C. Williams found four key areas of bias: the tightrope, prove it again, tug of war and the maternal wall bias (for women and caregivers). Backed by research from over 14,000 participants, she found that people in the majority group (white men) tend to believe the workplace is fair an average of nearly 20% points higher than folks from underrepresented groups (women and people of color).

Here are some areas of bias to look out for:

• The tightrope: For women, this could be being labeled as too aggressive for similar behavior to male counterparts, or being too passive if they don't express enough confidence. The middle ground between those two perceived behaviors is difficult. Women are much more likely to get feedback in performance reviews about personality traits like being too aggressive or not confident enough. There's very little room for error. This also happens at a higher level with other underrepresented groups.
• Prove it again: Underrepresented groups are often tasked with not just having to prove themselves once, but often multiple times to be viewed as capable as the majority group. This is often expressed in seemingly harmless biased statements like "he's a go-getter, he can figure it out," even though he's never done it before, or, "I don't know, she did it once, but can she do it again?" (even though she has performed before).
• Tug of war: Sometimes called the "queen bee" syndrome, underrepresented groups are often pitted against one another especially when there are few represented. This "zero-sum" game creates a perception that there is only room for so many people of color or women in a group, leading to in-group conflict.
• Maternal wall: This one affects all caregivers, especially women, even if they are not primary caregivers or do not have children. The Harvard Implicit Association Test reinforces this with 95% of people associating women with caregiving and men with providing. McKinsey cites the "broken rung" of fewer promotions and pay increases during the childbearing years for women and caregivers.

Affinity bias also affects most humans in the workplace. Even though we may believe that we want to be around different types of people, most people feel more comfortable around people like them. Sometimes called "mini-me" or "like me" bias, managers and people alike gravitate towards people with whom they can relate, rather than people different from themselves. This could be hobbies, gender, race, ethnicity, cultural background, class or other areas. Because managers are overwhelming of the majority group, this bias limits diversity and leads to promotion, pay raises and increased visibility for folks from underrepresented groups.

White men are more likely to be judged based on potential vs. actual performance. Separating these two terms is critical to inclusive performance reviews. Asking yourself, "how do I know that is true?" or "based on what information?" helps reduce this bias and focus on actual facts and real performance, not what is possible in the future. Most humans have seen white male leaders far more than underrepresented groups, and if not challenged, people can believe that is what leadership looks like. Challenging our biases helps us get more out of our collective talent and have a more inclusive workplace.

Related: Why Your Performance Reviews Are (Probably) Biased

What to do when you spot bias

Bias affects most people most of the time, yet there are concrete strategies to intervene and slow bias down. One of my favorite tools is Kristen Pressner's Flip it to Test it. To use her tool, consider the next time someone says something that's potentially problematic or biased. You get that pit of your stomach feeling like it doesn't feel fair and are not sure what to say or do.

Most people choose not to say anything, and therefore they let the behavior continue. Instead, especially in a performance review, flip it to test it. By asking, "would you say the same thing if a person were a different gender, race, ethnicity, ability or sexual orientation?" If the answer is yes, probably not. If the answer is no, there might be an issue with bias that needs to be addressed.

Inclusive leaders give feedback and "stretch" assignments equitably. Because most people feel comfortable with people like them, they are also less likely to share candid feedback with different people for fear of offending or not getting the message across correctly. To combat this, make sure feedback is a part of all one-on-ones throughout the year with all team members.

Performance reviews are not the time for surprises. Challenging or "stretch" assignments (temporary leadership, critical project to business, high visibility roles) are also often discussed at performance review time. And, these opportunities are more frequently given to the majority group. Pause and notice "who is not being included?" before making a decision about "stretch" assignments helps to create more equity.

Related: Entrepreneurs Should Watch Out for Cognitive Biases and the Curse of Knowledge

How to rid bias systemically

For inclusive leaders to thrive, bias needs to be rooted out of systems like performance management. One way to do this is to ensure goals are set and measured objectively. Oftentimes, goals are subjective and measured by opinion rather than actual results. Make sure all goals have a quantitative component with a clear understanding of what success looks like. If you're able to say "well, they tried" or "it was outside of their control," the goal was not objective and is inviting bias. By removing subjectivity from the process, you also remove human error and bias. Rather than managers promoting or giving pay increases to people they like, make sure it's based on actual performance.

Another way to eliminate bias at a systems level and not have to rely on managers to manage their biases themselves is to have performance calibrations. By inviting other perspectives into the process, performance is more holistically assessed across a team versus from only one individual perspective. Calibrations should not be a divide and conquer situation where talent is pitted against each other, and the focus should be on both actual performance and behavior to make sure that the team is rewarding the behaviors that get results.

Inclusive cultures know actions speak louder than words. Those that practice inclusive performance management keep their radars up for these biases, interrupt them when they see them and work to eliminate biases in performance reviews systematically.