top of page

Unveiling Hidden Biases: How Unconscious Bias May Be Impacting Your Organisation

The workplace now consists of more diversity than it arguably has before. With this comes a variety of lived experiences, cultural background and social perspectives. But despite the abundance of diversity around us, this can often be overshadowed by stereotypes, which in turn affects our attitudes and behaviours.

 

Sometimes these assumptions can be implicit and this is often known as “unconscious bias”. The EW Group, the UK’s leading diversity consultancy, notes that unconscious bias can often be “displayed subtly and without premeditation or intention” through our body language, mannerisms and dialogue.

In recent years, businesses have become increasingly conscious of and open to embracing diversity and inclusion in the workplace. But, there is still plenty of room for improvement. To prevent individuals from being subjected to unfair discrimination, it is important to highlight the ways that unconscious bias can manifest in your workplace, and recognise the patterns that cause it.

Here are a few critical examples of where unconscious bias can take place:

1. Biased hiring and exit processes

Unconscious bias can significantly affect a workplace’s hiring and exit processes. Starting from the very beginning of the screening process, recruiters or hiring managers may be influenced by preconceived notions and assumptions about a candidate’s prior experiences, their name, or their educational or work background. Applications can be rejected on the bias of the employer rather than the skill of the candidate.

 

A study launched by the British Academy revealed that racially minoritised applicants or those with non-English names have to send on average 60 per cent more applications to get a positive response from an employer than a white person of British origin. This is an example of name bias, where names act as racial indicators, and thus trigger a bias the employer may have toward people of those backgrounds.

 

Research conducted by The Leadership Quarterly — an international journal of Political, Social and Behavioural Science — found that even with identical qualifications to their white counterparts, ethnic minorities will receive fewer interviews or positive responses from employers — especially for those applying for a leadership position. This form of racial discrimination often hinges on the individuals’ racial perceptions of these names.

 

Similarly, unconscious bias can lead to overlooking a qualified employee's contributions or even holding them to a higher standard than colleagues. If someone is from a different background, they might even be viewed as a "troublemaker" for simply having a different communication style, leading to unfair discipline or early termination.

2. AI-influenced decision making

The growing use of artificial intelligence (AI) has reached unprecedented levels across all types of organisations and workplaces. Many workforces now use AI to help assist in the interview process and improve decision-making. But, AI is not without its flaws. It is designed, developed and deployed by humans who may have unconscious biases, and AI can perpetuate and propagate these biases.

 

As a result, the systems may generate outputs that are discriminatory, misleading, harmful, or even illegal. While eliminating bias is no easy feat, to mitigate AI bias, we must be aware of the ways discriminatory data can be introduced into the AI systems at multiple stages of the AI lifecycle:

 

●        Data collection and labelling: The data used to train and test AI systems may not be representative of the target population, or may be labelled by humans who have implicit biases. This can result in the exclusion or underrepresentation of certain groups, human errors, or labels that are inconsistent, subjective and reflect human judgements – which could eliminate qualified job applicants. For example, training data for facial recognition, which is developed mainly using the faces of white people, can negatively impact recognition accuracy for people of colour and lead to higher error rates.

 

●        Algorithm design and testing: The problem with many data-processing algorithms is that sometimes they are not tested thoroughly enough to detect and correct potential biases. They may be tested on data sets that are not representative of real-world scenarios. For instance, AI models have increasingly been used in medicine to diagnose specific diseases or evaluate X-rays. However, if they are predominantly trained using data from white patients, this can lead to inaccuracies when applied to ethnic minorities. Even if a healthcare provider removed the AI guidance, the biases absorbed could still influence decision-making in the future.

3. Microaggressions in virtual spaces

When striving for an all-inclusive workplace, it’s important that employees feel a sense of belonging, and employers foster an open and honest safe space. However, this can be infringed upon by microaggressions.

 

Microaggressions are defined as “thinly veiled, everyday instances of racism, homophobia, sexism.” Sometimes it can be disguised as an insult, other times it's an offensive comment or inappropriate gesture.

 

According to the Deloitte State of Inclusion report, it was found that 83% of respondents were exposed and/or witnessed this form of discrimination in the workplace, including both subtle and indirect comments. Furthermore, since the COVID-19 pandemic, many workforces moved to virtual work settings and this resulted in a surge of microaggressions.

 

Microaggressions can manifest in virtual spaces and range from blatant to subtle insults including:

 

●      Microassault: an explicit or nonverbal action intended to hurt the victim through name-calling, racial slurs, avoidant or discriminatory behaviour.

●      Microsinsult: a subtle snub that can belittle a person’s heritage, identity or background.

●      Microinvalidation: comments which negate the feelings or experiences of marginalised individuals.

 

Many DEI experts recommend action-focused training sessions for your workforce, teaching the ability to recognise these types of microaggressions; which will ultimately result in creating a more inclusive online space.

Maximising the impact of unconscious bias training

By increasing our awareness of unconscious bias, we can begin to make a difference in the workplace and implement measures against it.

 

It’s important to be open to discussion and having challenging conversations with your colleagues. Unconscious bias training can come in particularly useful if a team is struggling to promote inclusivity or create a safe space. It can lead to a fairer and more inclusive workplace and also make people aware of potentially harmful unconscious biases; which can reduce the impact of those biases when interacting with others.

 

Unconscious bias training is designed to be enlightening, interactive and cross-cultural. Although some remain sceptical, there is conclusive scientific evidence that unconscious bias training does work. Several bodies, including the Equality and Human Rights Commission in 2018, reported that training interventions were effective in reducing unconscious and implicit bias and that long-term training sessions were necessary to drive long-lasting change.

 

Training programmes can affect your organisation for the better. Sessions will leave delegates acknowledging the subtle ways unconscious bias can play out at work, how various aspects of their role may be open to unconscious bias, identify opportunities where they can promote the best practices and employ a set of practical tools or action plans that can assist with addressing unconscious bias in the workplace.

コメント


bottom of page