Culture shock: Are psychometric tests discriminatory?
According to the Equality Act 2010, businesses are legally barred from discriminating based on age, disability, gender, race, religion, or belief. But the need to avoid bias goes beyond simple legislation — but the smart business decision: according to a LinkedIn study, diverse companies earn 2.5x higher cash flow per employee, are 35% more productive, and make better decisions 87% of the time.
One of the ways organisations are trying to combat bias and discrimination in their hiring, is by employing a more robust, data-based selection process. By minimising ‘hunches’ and gut-feelings, and instead relying on tangible information in recruitment, companies can be assured that they are hiring the best person for the role rather than someone who looks like them or fits a stereotype.
Psychometric evaluations — assessments that allow employers to understand their candidates’ personality and cognitive abilities — are a favourite among hiring managers for this reason. However, as an employment gate-keeper, many have questioned whether, rather than helping organisations avoid discrimination, they create bias in-and-of-themselves. So, it’s time to ask, once and for all — are psychometric tests discriminatory?
What is bias?
In hiring, bias refers to the act of discriminating against a certain group of people based on preconceived notions regarding their abilities. Bias can be either conscious — for instance, choosing to hire a man over a woman because of the concern she might get pregnant and require leave — or unconscious — not hiring a fully qualified woman for a senior position out of a gut-feeling that women are not as good at managing people.
To put it plainly, bias in recruitment would cause you to hire someone not because they are truly the best for the role and the company, but because they fit into a box of ‘the kind of person’ who’d do the job well which is based on their class, gender, ethnicity, nationality, age, religion, or culture.
It’s important to note that there isn’t anyone who’s completely free of bias — we all have certain ideas and beliefs that could bias our decisions. This is why it’s important to consciously try to mitigate their impact in a selection process.
Further reading: How to combat unconscious bias in hiring
What kind of bias can psychometric assessments promote?
Psychometric assessments are a scientific method used to measure individuals’ personalities and cognitive abilities. Businesses use them as a relatively objective data point to quickly shortlist large pools of candidates, or to gain more insight into candidates. However, as psychometric tools are designed and developed by humans, they are not by definition free of error or bias themselves.
Cultural discrimination in psychometric assessments
Psychometric assessments can reproduce bias in a number of ways:
1. Language
Language can be a serious barrier to individuals’ ability to complete tests properly. For example, if your first language isn’t the language in which you’re taking the assessment, you might struggle to understand certain phrases or directions as quickly and effectively as a native speaker. This is particularly true in cognitive tests for verbal reasoning, but can impact every type of psychometric assessment. Even if a translation is provided, literal interpretations or mistranslations can impact the score.
2. Question design
For candidates from a diverse cultural background, certain scenarios may be irrelevant — winter sports for someone who comes from a tropical climate, for example. What’s more, different cultures have varying values, norms, and moral considerations, making their responses depend on the ‘baggage’ they come with. If the assessment is designed without appropriate reflection on these differences, it might unfairly score those who come from different backgrounds to the test designers.
Another element that could play a role in psychometric discrimination is how familiar the candidates are with taking similar assessments. It’s likely that an applicant from a Western country who went to university would have many opportunities to respond to similar tests, and would be able to do so with relative ease. Access to education can also impact results if assessments aren’t designed appropriately.
Different cultures might also have different ways in which they answer questions — some may be more likely to exaggerate, while others encourage modesty. Without being able to properly contend with these human differences, assessments can be biased.
3. Norm groups
Perhaps most importantly, psychometric tests rely on norm groups to be able to score people. This means that the assessment uses results from multiple people to decide what’s considered a high or a low score. If this norm group — the collection of people upon whom the results rely — don’t come from a diverse background, assessments run the risk of not being applicable to everyone, but only to a specific group based on gender, culture, or other factors.
Neurodiversity discrimination in psychometric assessments
The human brain is fantastically diverse — this means that each of us is different. However, certain groups of people have an especially unique way of thinking, often referred to as neurodivergence. This includes autism, ADHD, and dyslexia, for example. Because psychometric assessments are usually normed based on neurotypical people, candidates with certain neurodivergence conditions might score lower in them, not because they are worse but because the tests themselves aren’t structured to accommodate their distinctive way of thinking. For instance, neurodivergent candidates may be more anxious, suffer from executive function challenges, or over-analyse questions.
This is especially critical as recent cases, such as British Telecommunications vs Meier, highlight the requirement of businesses to proactively accommodate and offer reasonable adjustments for neurodivergent candidates when taking psychometric assessments.
Further reading: How to ensure neurodiversity in hiring
How do psychometric assessments combat bias?
While psychometric assessments can pose many challenges in terms of bias, they are not inherently biased tools. In other words, they are just as biased as their designers make them!
Psychometric assessments have the potential to be extremely objective, unbiased tools, that can even help businesses mitigate bias in the selection process. When formulated correctly, these tools can be highly valid (measures what it claims to measure) and reliable (reproduce a result consistently), which means they can help predict performance with minimal or no bias.
Some of the methods well-designed psychometric assessments fight bias include:
- Involving diverse cultural groups in both the development of the assessment and the norm group
- Providing tests in multiple languages with high-quality translations
- Consistently re-testing and re-norming assessments
- Suggesting reasonable accommodations for neurodivergent candidates, such as extra time or open-ended questions
When purchasing a psychometric assessment for your business, it is vital that you question your provider about the science behind their tests and the methods they use to eliminate bias in their test design.
Further reading: What you need to know before buying psychometrics
How to avoid discrimination with psychometric tests
As we’ve mentioned above, a scientifically-robust psychometric assessment can be a relatively objective tool to help with your selection process. This means it not only would not induce bias, but help you fight the unconscious bias we’re all likely to have during interviews.
If your hiring process is done correctly, psychometric assessments would be one of a few factors you take into account when selecting a candidate. It can inform your interviews, which should be structured and standardised, to gain useful data to help your decision, instead of relying on hunches or gut-feelings which are usually filled with bias.
If you want to learn more about how data and psychometrics can help you mitigate bias in hiring, book a demo with us today.