The two terms "Conscious Bias" and "Unconscious Bias" are creating a lot of buzz. What do they really mean? The former is logic and reason driven while the latter exists outside of our conscious awareness. Family, cultural conditioning, inherited belief systems and memories can have short-term and long-term impacts on our conscious awareness. Many biases are formed and shaped throughout a human lifespan and are held at a subconscious level, primarily through societal and parental conditioning. If you hear the word “kitchen”, do you instinctively associate a gender with it – masculine or feminine? In many gendered European languages including German, Italian, French and Spanish, the word “kitchen” and words for most kitchen appliances are linked to articles used to describe the feminine gender, which shows a deep-rooted bias in languages. Even seemingly unbiased words such as ‘forefathers’, ‘mankind’ and ‘mother tongue’ exhibit inherent gender bias in our society. However, most of us would not even notice the subtly hidden stereotypes in these words unless we make a conscious effort to acknowledge biased representation. These are simple examples, but we need to ponder it over a bit more to understand the root of it all.
Unconscious or implicit biases are far and wide, from the locality that we choose to live in, the close association that we make and the people we choose to date. Though most of us have difficulty accepting or acknowledging it we all do it. We are social beings who prefer the company of other humans. We organize ourselves into various kinds of social groups based on our preferences and inclinations. The past few decades of neuroscience and behavioural research has led to an immense wealth of information explaining the whys and hows behind unconscious bias. Prof. Dr. David M. Amodio wrote in his research paper published in Nature Reviews Neuroscience, 2014, that the human brain has the natural tendency to sort people into groups, differentiating ‘us’ from ‘them’. The neuroscience of prejudice and stereotyping is rooted in the amygdala, the unconscious mind that helps us process billions of stimuli on any given day. A human brain gathers trillions of bytes of information and processes them in a certain way – unconsciously categorizing, formatting and transferring them into familiar patterns. Our brains are able to quickly decipher which information we should focus on. We can use this information to survive, make assumptions and inferences and feel emotions that cause us to be attracted to certain people but not to others.
Historically, humans have developed a fundamental ability to organize the social world around them into categories. This ability to attach “stereotypes” to social groups, be it racial, gender, age, body size, profession, sexual orientation or religious beliefs, has helped us to navigate the world without being overwhelmed by information. The downside is that this is hardwired in our brains and it could be challenging to be aware of it, let alone change it. Since the 1960s behavioral scientists and psychologists have been researching these cognitive biases. Affinity bias, attribution bias, ageism, beauty bias, conformity bias, confirmation bias, contrast effect, name bias, gender bias, halo/horns and weight bias effect are the most common types of biases existing in our times.
While unconscious biases are natural and unintentional, they should not be left unnoticed and uninterrupted. In the age of machine learning and artificial intelligence (AI), when human programmers are designing algorithms, we need to address unconscious bias more than ever. No one wants Google and Facebook AIs to be thinking that 90% of professors in the USA are white men. ImageNet Roulette, a project executed by researcher Kate Crawford and artist Trevor Paglen has unraveled the dangers of feeding flawed data into AI through their studies on ImageNet, which is a database of millions of images and is a very popular training set used for machine learning. This project demonstrated that the best AI vision system might see a picture of your face and also detect your race, a gender or a term that criticizes and challenges your character and intentions. ImageNet eventually accepted the flaws in their database and have been working to fix them. Different types of biases have been identified in facial recognition technologies, recruiting programs, the algorithms behind dating apps and various web search engines. If we do not acknowledge the problem and allow it to remain unnoticed it can stunt innovation and productivity. It is necessary to understand the different aspects of our lives from standpoints such as our socio-political-economic environment, our own personal experiences, cultural context and social conditioning, to the attitudes and stereotypes that have developed over time leading to faulty decisions and incorrect conclusions based on feeble logic. It is important to remember that unconscious biases may be incompatible with an individual’s social values.
Unconscious biases also lead to devaluating extraordinary talents from the job market. A study conducted in 2016 by two researchers, Dr. Eva Zschirnt & Dr. Didier Ruedin, showed that ethnic and racial discrimination has remained dominant in the hiring process despite the introduction of anti-discrimination legislation across OECD countries in the last 25 years. This study also reported that minority candidates with comparable CVs needed to send over 50% more applications to receive an interview call than candidates belonging to majority groups. A study showed that immigrant groups, especially those originating from non‐European countries, tend to experience disadvantages in the labour market and have longer periods of unemployment due to migration‐related factors including discrimination by employers.
Many social experiments display deeply rooted gender bias, favouritism and pay gaps in STEM that are pervasive and have remained completely unacknowledged until recently. A 'pay equity analysis' performed by Payscale (2020) showed that organizations pay women less than men for equal work. Moreover, in some states in the US, the gender pay gap is wider for women of colour, women in executive level roles and women in certain occupations and industries. According to the gender pay gap statistics by Eurostat (2018), women's gross hourly earnings were on average 14.8 % below those of men in the EU. In this study, the highest gender pay gap was recorded in Estonia (22.7 %) and the lowest in Romania (3.0 %) in the European Union. A study led by Prof. Cristian L. Dezsö and Prof. David G. Ross (2012) showed that women’s representation in top management brings in informational and social diversity to the organization. It also improves the firm’s performance.
To counteract unconscious bias, a number of strategies including psychometric tests, training programs, short exercises before making significant decisions and meditation have been proposed. Psychometric assessments such as the Implicit Association Test introduced by Greenwald, McGhee & Schwartz (1998) are used in psychology research to detect unconscious bias. These tests are able to point out areas of unconscious bias that one can consciously work on. Nowadays, firms such as Facebook and Google provide unconscious bias training to all employees. Such training programs are helpful with developing the sociological counter narrative and provide tools to rectify unconscious patterns of thinking and eventually discontinuing discriminatory conduct. The workshop introduces bias-busting techniques which can help alleviate the potentially negative influence of unconscious bias within an organization.
Diversity and inclusivity help to tackle unconscious bias. It is not difficult to imagine that a person with a lack of exposure to diverse social groups is prone to unconscious bias. Research has shown that diverse groups can be better at making decisions than homogenous ones. Diversity and inclusivity confer a positive impact on any organisation and often these companies turn out to be the most successful ones.
The introduction of blind hiring has a dramatic impact on talent selection where personal information is stripped from the resume and recruiters and managers evaluate the candidates only on the basis of qualifications relevant to the position requiring to be filled. Single or double-blind evaluations conceal the identity of the candidates or their gender, race, position and institutions of education. Statistically, it increases the chances for a woman to advance to the next stage. The extremely popular TV show “The Voice” is inspired by blind auditions to encourage unbiased selections. The study "Orchestrating Impartiality: The Impact of 'Blind' Auditions on Female Musicians" conducted by two researchers, Claudia Goldin and Cecilia Rouse (1997) was a great revelation of gender-biased hiring. It showed that switching to blind auditions led to an increase of 30-55% in the proportion women amongst new hires. Spark, a new funding scheme of the Swiss National Science Foundation (SNSF) started in 2019 is one of its kind in awarding grants to researchers solely based on anonymous research proposals to circumvent any kind of bias, ensuring that evaluators assess the project and not the applicants. Blind recruitments can help the recruiters overcome some, if not all, of the unconscious biases.
Unconscious biases are built in our psyche. On several occasions, one has to make judgments in a fraction of a second well before the critical thinking part of our brain kicks in. Even when we say that we are open-minded and impartial these biases still permeate our minds and at times these mental shortcuts can lead us to misjudge people. The knowledge of the existence of unconscious bias is more than two generations old. However, we as a society have not progressed much, owing to reasons such as lack of acknowledgment and action, which do not always go hand in hand either.
So, what can be done to prevent unconscious biases? Firstly, we need to acknowledge that being human we all have our own unique unconscious biases. There is no need to be ashamed of that. We can then look for these biases within ourselves. There are several tests and questionnaires to help us identify them. Only when we are aware of our prejudices, can we actively mitigate them. Once each-and-every one of us is able to bring this into practice in our daily way of life there is hope that the society will implicitly be free of unconscious bias.
Dr. Vandana S. Kushwaha
Scientist, Biophysicist