Common unconscious biases in performance reviews – and how to address them

Learn how to counter the common types of unconscious biases in performance reviews.

Common unconscious biases in performance reviews – and how to address them

Performance reviews are used in virtually all organisations to set remuneration, make promotion decisions and for L&D planning. They aren’t immune, however, to human psychology.

What is unconscious bias?

Unconscious biases are quick, instinctual judgements we make about a person or situation. All humans have biases. They are shaped by our beliefs and past experiences. It’s the brain’s way of making sense of things when we are bombarded by information and constant stimuli. These cognitive shortcuts may have been useful once upon a time when our ancestors were facing down sabre-tooth cats, but in the modern workplace, unconscious biases can cause us to treat people unfairly and perpetuate stereotypes that harm diversity and inclusion.

In this blog, I’ll cover some common unconscious biases in performance reviews. I’ve experienced many of them in my career (both as the manager and the employee). I’ll then share some thoughts on how to overcome them. 

Types of unconscious biases

Infographic showing common unconscious biases in performance reviews. 1. Affinity bias [image of two similar-looking men high-fiving each other.] 2. Gender bias [image of weighing scale with the male and female symbol on each side. The female symbol weighs less.] 3. Recency bias. [image of hourglass with sand running down.] 4. Central tendency bias. [image of circles surrounding a larger circle in the middle. There are arrows pointing from each smaller circle in the outer ring to the inner circle.] 5. Halo effect bias. [Image of an angle's halo.] 6. Horns effect bias. [Image of two devil's horns.]
Common types of unconscious biases in performance reviews.

1. Affinity bias

What is affinity bias?

Affinity bias occurs when we favour people from a similar demographic or background or with a similar personality or interests. A manager might unconsciously rate an engineer with whom they share a strong rapport more highly. 

Employees who don’t share the same affinity with their leadership may feel undervalued and demotivated about their ability to advance. Left to fester, this could stifle innovation in the workplace as people who think in similar ways are less likely to challenge the prevailing way of doing things or offer new ways of solving problems. 

How managers can address this:

The key to addressing affinity bias is implementing a standardised system for evaluations that include clear, specific criteria relevant to the job performance. Using the same framework to evaluate all employees minimises the role of subjectivity and gut feel in assessing performance. 

Managers should recognise that diversity in thought, perspectives and experiences adds value to the team. Work with other managers and team leads to build a more inclusive culture. For example, look at mentorship and sponsorship opportunities that are open to all team members, rather than just letting these form through natural affinities, since the latter is more likely to benefit those in the ‘in-group’.  

What to do if team members experience this from their manager:

If you sense that affinity bias is clouding your manager’s judgement, request feedback from them based on your specific achievements and ask for clarification from your manager on how performance metrics are applied fairly across the team. 

It may be tempting to withdraw if you and your manager don’t get on like a house on fire, but it’s important to keep sharing your progress with your manager and keep your achievements visible. You don’t need to be best friends with your boss, but you need to work with them to obtain actionable feedback and set goals to help you grow. 

2. Gender bias

What it is gender bias?

Gender bias occurs when managers allow their assumptions about gender to unfairly influence their evaluation of their team members’ work. Sadly, this is one that my Kaleida co-founders and I have grappled with throughout our careers.

For example, there are persistent stereotypes about women being ‘less technical’ than men or being more suited to front-end development, while back-end or full-stack development is seen as the domain of male software engineers. (This stereotype ignores how much more complex and nuanced front-end development has become over the years – but that’s the subject of another blog post!) 

Additionally, men are more likely to receive concrete feedback related to their performance, while women tend to get more vague feedback on their personal attributes, such as their communication style or likeability. 

How managers can address this:

Using a standardised career growth framework and building a real-time feedback culture can help reduce gender bias. Studies have shown that poorly defined, ‘open box’ performance review processes allow gender biases to creep in. When managers don’t have a clear sense of what they are supposed to be evaluating, they start to fall back on deeply ingrained cultural assumptions about how men and women are ‘supposed’ to behave. 

What to do if team members experience this from their manager:

Have clear documentation of your contributions and achievements. That way, if you feel you are getting vague feedback from your manager and suspect that gender bias is playing a part, you can redirect the conversation to focus on facts and evidence and ask for specific advice on how to improve to get to the next level.

Solicit feedback from peers and leaders in the org who can give you feedback and testimonials on your performance in a particular project. This provides a more comprehensive and holistic view of your impact to your manager, who may otherwise attribute this to communal effort rather than your individual accomplishments. 

Thumbnail image with text 'how to get promoted to senior software engineer'. Features an image of a hand holding up blocks that are stacked to look like a staircase. A pen scribble bounces from one step up to the next, signifying career progression.

READ OUR BLOG POST: How to get promoted to senior software engineer

3. Recency bias

What it is recency bias?

Regency bias is the tendency for managers to emphasise the most recent events when evaluating performance. For example, an engineer who had a less-than-average performance all year but delivered a successful project in the quarter before the performance review period may be rated higher than one who contributes consistently all year but may have had a few blips with a challenging assignment in the period immediately before review time.

How managers can address this:

It’s human to revert to the events that are fresh in our memory, but it doesn’t tell the whole story. Managers should work with team members to record all their achievements throughout the year. Make a habit of noting down your team members’ contributions. Regularly solicit feedback from stakeholders on your team after they have completed major milestones and projects throughout the year, so you have a comprehensive list of evidence to draw on when conducting performance reviews. 

What to do if team members experience this from their manager:

If you proactively maintain and surface a record of your accomplishments to your manager in your regular one-to-ones, you will find it easier to prevent recency bias from affecting their appraisals. If you find your manager focusing on actions that occur just before the performance review, remind them of other projects you worked on in the entire evaluation period with your detailed log. 

Managers often resort to recent events when they are time-poor and don’t invest enough time to recall and evaluate your performance thoroughly. Make it easier for them. Ahead of the review meeting, send your manager a comprehensive self-evaluation of what you achieved, what you have learnt, and how you have further developed and strengthened your skills since.

4. Central tendency bias

What it is central tendency bias?

Central tendency is a ‘play it safe’ phenomenon where managers rate most employees as average to avoid extremes. High-performing software engineers might receive a "meets expectations" rating because the manager is wary of using "exceeds expectations" for anyone. This also affects underperforming software engineers who may have their performance rating artificially inflated by being skewed to the middle of the pack. This means that top performers aren’t recognised and rewarded for exceptional work, while underperformers miss out on feedback to help them improve.

How managers can address this:

Central tendency bias occurs when managers lack the confidence to differentiate between different performance levels. A lack of clarity on how to evaluate performance can cause managers to revert to middle-ground ratings. 

A career growth framework with clearly defined criteria for performance at each level can solve this. Alongside an engineering growth framework, ask for training to develop your evaluation skills to recognise the full spectrum of employee performance. 

What to do if team members experience this from their manager:

If you feel that your performance is not accurately reflected due to central tendency bias, ask them to review a detailed record of your accomplishments. Sometimes, managers default to average ratings if they don’t have enough visibility into your work.

If you are getting average ratings consistently across the board even though you feel you have excelled in some areas, ask your manager to spend some time with you discussing the specifics. For example, if you are getting excellent feedback from peers about the quality of your code reviews but are getting average scores from your manager, say: 

“I noticed that you listed my code reviews as only meeting expectations. I get lots of great feedback from the junior engineers about how thorough and helpful my reviews are. I take the time to consider how the code works with the broader architecture, provide clear and specific feedback about the changes I am asking for, and order my requested changes by priority level. Can you show me examples of what you consider a great code review so I can improve?” 

By seeking precise feedback and sharing your work, you can get a clearer picture of your performance and even ask for a reassessment if the initial rating is inaccurate.

5. Halo and horns effect bias

What is the halo and horns effect bias?

The halo effect is when one positive aspect overshadows all other areas of an individual’s performance, while the horns effect bias is the opposite, where a single negative incident can affect the overall review. 

An engineer who created a groundbreaking feature may receive high ratings across the board. The manager may make excuses for them even if they were not timely in their communication or if they introduced severe errors into the code base during the project. Conversely, another engineer’s minor slip-up on a project could overshadow an otherwise stellar and consistent performance.

Managers who have formed an early negative impression of one of their engineers may only notice errors and overlook successes, or vice versa - a form of confirmation bias. Often, the perceptions are so deeply rooted that they persist for the entire working relationship, hampering the team member’s ability to progress under their leadership. 

How managers can address this:

Managers first need to strive to be aware of this bias. When considering each team member and their overall impressions of them, managers should challenge themselves to look for specific evidence that substantiates their impressions. Rather than focusing on personal attributes, evaluate the employee on measurable and observable criteria - and ensure you use the same criteria for all team members.

Collecting data from different sources can mitigate the tendency to discard evidence that doesn’t align with your point of view. Seeking 360-degree feedback from other managers and the team members' peers can bring in new perspectives you may not have considered. 

What to do if team members experience this from their manager:

If your manager’s feedback keeps returning to one or two isolated incidents, they may carry some of this bias. Ask for specific feedback for various areas of your work so your manager can evaluate your performance more comprehensively. If there were mistakes you made early on, discuss how you've learned from the error and provide evidence of your growth over time. 

Conversely, if you think you might benefit from the halo effect, it’s tempting not to do anything. Recognise, though, that it can harm your career growth if this leads to your complacency or invokes resentment among your peers. Ask your manager to specify what you have done well, and then work with them to identify your growth areas. Look for opportunities to take on stretch goals to continue improving.

How career frameworks can help overcome unconscious biases in performance reviews

Looking back at the unconscious biases I’ve outlined, we can see a consistent pattern. All unconscious biases have something in common. They arise when there is an absence of data since they rely on impressions, gut instinct and subjective interpretations of events. 

A powerful tool to overcome unconscious bias in performance reviews is to use a standardised career growth framework like Kaleida in your engineering organisation.

By clearly outlining the competencies, skills and achievements software engineers should have at each level, organisations can ensure an objective benchmark against which all employees can be assessed. This helps keep managers accountable for evaluating their team members’ work based on specific and measurable criteria instead of assumptions or personal preferences - the stuff that biases feed off.

Screenshot of Kaleida career growth platform showcasing a heat map of team member skills and a spider graph showing collated strengths across a team  10 key skills - agile, coding, craft, DevOps, security, coaching, communication, community recruiting and wellbeing."

Career frameworks boost motivation and performance by helping team members understand what is required for their career progression. It inspires confidence that their performances are being evaluated fairly. As a result, you’re more likely to retain great engineering talent and build a thriving, high-performance culture that rewards excellence.

Are you ready to eliminate unconscious bias from your organisation and start implementing effective and high-impact performance reviews? Get in touch with Kaleida today.