'Invisible Women': 10 ways gender biased data governs your daily routine
Data bias at the centre of gender discrimination
Fri 12 March 2021
What is the meaning of gender bias and how does it affect our every day?
Gender bias towards men can be seen everywhere; we know that a cartoon hedgehog is male because he is normal – he does not wear a pink bow and simper. We also only get a substantial number of women in orchestras when we utilise ‘blind auditions’.
On average, women are more educated than men. For decades, women have earned more bachelor’s degrees, master’s degrees, and doctorate degrees than men. Yet women still earn less than their male counterparts. A university education leads to an increase in net earnings of about 20%. But the difference between male graduates and female graduates is stark: ‘the estimated gain to the exchequer of individuals attending HE is around £110k per student for men and £30k per student for women’.
Caroline Criado Perez won the 2019 Financial Times and McKinsey Business Book of the Year Award for Invisible Women . Her main argument is that ‘most of recorded human history is one big data gap ... the lives of men have been taken to represent those of humans overall’, and that designers and developers have perpetuated this bias towards men in the data they use.
Gender Bias Examples
- Modern devices don't come in women sizes. Phones are built with male body dimensions in mind and are often too big for female hands. After the launch of the 2018 iPhone X, female customers criticised Apple for creating a phone that was too large for the average woman’s hand, making life frustrating for many and giving others repetitive strain injury.
- Reaching for cereal in the morning can be a similar struggle. Shelves are usually set to the standard male height, making them a bit of a stretch for the average woman.
- Women's clothing tends to have smaller pockets - if any at all - than men's. There is no supporting data set that found that women don't need or use pockets. Rather, this is because throughout history it was not considered fashionable for women to have large pockets: 'men are busy doing things; women are busy being looked at. Who needs pockets?' [...] Men's dress is designed for utility; women's dress is designed for beauty'.
- If you are a woman and ever get annoyed at speech recognition technology mishearing what you say, it may be because technology like Google Home is 70% more likely to recognise male speech. Speech-recognition technology is trained on large databases of voice audios, that are dominated by male recordings.
- Women are 47% more likely to be seriously injured in a car crash. Most crash-test dummies are modelled after a man’s average weight and height, so women have a higher chance of severely injuring themselves than men in comparable collisions. Cars are also designed according to the male anatomy: women tend to sit further forward than men when driving because, on average, they are shorter. Women’s legs need to be closer to the pedals, and they need to sit more upright to see clearly over the dashboard, but this isn’t the ‘standard seating position’.
- Settings for heating and air-conditioning systems are adjusted to men’s higher metabolic rate which might be why women find their office colder than men. This can make it difficult to concentrate and lead to lower productivity in the workplace. Although offices are closed to most people during the pandemic, women who are currently working in the medical sector are still suffering from data bias. Many women have complained that PPE is made to fit a ‘6ft 3in rugby player’, some female staff are reported to have developed ulcers on their faces from ill-fitting PPE, others have to roll up the sleeves of their fluid-repellent gowns and some have to use micro tape to seal gaps around their jawline. Health professionals, experts and unions say that poorly fitting equipment is risking the lives of female workers. Dr Helen Fidler, the deputy chair of the British Medical Association (BMA) UK consultants committee, said ‘We know that properly fitted PPE works, but masks are designed for a male template, with the irony being that 75% of workers in the NHS are female.’
- The design of cities can be a big problem for women. Cities have been laid out ‘to serve the needs of this mythical male breadwinner who has a wife at home in the suburbs’. The man drives to work and then home to a place of leisure, with an average twice daily commute. So there aren’t as many services accessible to women, in residential areas. But women’s travel patterns tend to be more layered, ‘they’ve got to take kids to the doctor, to school, get groceries, check in on a relative…’, they make lots of short interconnected trips, and are more likely to use public transport than men. Travel route data was finally taken into consideration ‘when local officials in the town of Karlskoga in Sweden looked at their snow-clearing schedules, and realized that they had designed them to meet the needs of men. As a result, the order in which the snow was being cleared (major roads first; local roads and sidewalks second) benefitted men. So they decided to switch the order around—and found to their surprise that the number of admissions to the emergency room fell dramatically. Because it wasn’t men in their cars who were falling over and fracturing their bones: it was women pushing buggies through the snow’.
- You may have noticed that the queue for women’s public toilets is often far longer than the men’s, and there’s a tendency to blame women for this rather than the male-biased design. On average women take up to 2.3 times longer than men in the bathroom, but ‘women make up the majority of the elderly and disabled, two groups that will tend to need more time in the toilet’. Women are also more likely to be accompanied by children or disabled and older people. There are also 20-25% of women of childbearing age who may be on their period at any one time, and therefore need time to change a tampon or a sanitary pad. Women in general may also require more trips to the bathroom; ‘pregnancy significantly reduces bladder capacity, and women are eight times more likely to suffer from urinary-tract infections than men which again increases the frequency with which a toilet visit is needed’. So the 50/50 floor space of male and female bathrooms is not as equal as you may think. Men have urinals so more men can pee at the same time, and it’s also more socially acceptable for men to pee outdoors than women. At festivals for example, men can save money and time whilst women lack access to toilets.
- For women going through a painful period, the story is often told of the accidental discovery of sildenafil (Viagra). Intended as a treatment for angina (chest pain), the clinical trial revealed a side-effect: ‘magnificent erections’. If the original trial included women, we might have discovered that sildenafil can also treat severe period pain, but alas, ‘men got their miracle drug’ and ‘women are still waiting’. As Perez reports in her book, we can’t confidently prescribe sildenafil as a treatment for period pain yet because ‘only a small and suggestive trial has yet been funded’.
- Women are 50% more likely to be misdiagnosed with a heart attack. Women often don’t present the same heart-attack symptoms that doctors are taught to look out for, namely chest and left-arm pain - typically experienced by men. Women, particularly young women, may in fact present without any chest pain at all, but rather with stomach pain, breathlessness, nausea and fatigue. Yet these symptoms are often referred to as ‘atypical’.
Unfortunately, when it comes to women of color, disabled women and working-class women, the data is practically nonexistent: ‘Not because it isn’t collected’, Perez explains, ‘but because it is not separated from the male data’.
How deliberate is this ignorance of the gender data gap?
One thing Perez points out in her book is that the gender data gap is not generally malicious, or even deliberate. ‘It is simply the product of a way of thinking that has been around for millennia and is therefore a kind of not thinking’.
In an interview with Wired, she was asked ‘at what point though—especially now that we have access to more data sets...does the ignorance of data become deliberate?’
She claims that sometimes women are forgotten by accident because a male-biased team has made a decision and forgotten that women exist. Other times, it might just be down to the fundamental difference between male and female human experiences.
Perez says that ‘the point where I start thinking about this as a political project is when you start getting to the excuses...at that point it’s not forgetting. It’s about excluding’
What are some serious consequences of gender bias psychology?
Gender bias exists in data because it exists in the real world.
The reason we need to be picky about the representation of women in data collection goes beyond being able to reach for a shelf or hold a phone comfortably. It reflects a wider problem that gender bias can have very serious implications in society.
In the UK, a woman is killed by a man every three days. Globally, six women are killed every hour by men and over 80% of murders are committed by men. Many are calling it a ‘global pandemic of femicide’. ‘It's the behaviours that we encourage in little boys and little girls, says domestic abuse campaigner Karen Ingala Smith. ‘Women are turned into products and men into consumers. And it's a consumer that has the power rather than what they're buying’.
In order to prevent the above-mentioned crimes the conversation and focus seems to always revolve around what clothes women should wear to better deter rapists, what times women can walk around alone at night, even what kind of hairstyle attracts rapists and how to behave around men in order not to give them the wrong signals.
When have you heard of little boys at school being taught how to manage their anger or aggression? Or how to manage their ‘desires’ in a situation in order not to rape someone? When have we told men to stop walking around at night, or set curfews on men, because they pose a danger to society? This societal and cultural problem is just one example of how gender bias is still present today.
How can we avoid gender bias in data research?
With all this said, there are some ways which can help you avoid data bias.
- Research your users in advance to the best of your ability
- Ensure your team of data scientists and data labellers is diverse, includes women, people of various ethnicities, races and economic backgrounds should be included.
- Combine inputs from multiple sources to ensure data diversity.
- Enlist a third-party to review your collected data to ensure no biases have been overlooked.
- Analyze your data regularly, keep track of emerging data and information as well as errors and problem areas so you can respond and resolve them quickly.
The gender data bias against women is shocking and present in everyday activities. Perez’ award-winning book Invisible Women points out various examples of this and leaves you wondering how such bias still exists today, not only in data but in the way we conceive of women’s position in society as a whole. Luckily, there is hope! The president of the Royal Statistical Society, the managing director of the International Monetary Fund and both heads of the UK Government Economic Service are all women.
Filling data gaps takes time, but with enough push from both men and women we could get there a little faster.