Opinions expressed by Entrepreneur contributors are their own.
Women are still considered second, and you might not even realize all the ways this is done. Some are obvious: Women earn less than men and there are fewer of them in leadership positions within organizations and on boards. Women’s career trajectories were also significantly sidetracked (or ended) as a result of the pandemic at a rate not seen with respect to men’s career paths. Women’s perspectives and experiences are downplayed, and these contributions to gender bias and gender gaps. Their needs are taken for granted and/or lumped into the bubble of male needs or wants.
One lesser-known way these disparities occurs is through data. You might be saying, “How can that be? Data is based on research and facts. How can there be a bias that data actually creates?“The answer involves a disregard for women in the research — from business to technology to medicine and other practical aspects of life.
What is the data bias?
Data is collected in order to provide evidence of what works (or doesn’t) for various projects, concepts or innovations. It allows researchers to know what needs to be adjusted and/or moved forward. One variable within research that makes investigation challenging is comprehensiveness: All aspects of the application of data need to be considered. When an aspect is overlooked, negative or dangerous outcomes can result. Consider, for example, the case in which a self-driving Uber car hit and killed a woman in Arizona. Uber determined that the car could not identify that an object was a pedestrian unless it was near a crosswalk — an oversight that created a dangerous data bias.
Also consider facial recognition software, a tool used by many law enforcement organizations. After it was being applied to identify, target and convict criminals, research revealed that in many cases its algorithms were significantly less accurate when it came to people of color as well as females. The resulting bias had widespread impacts on both civil rights and public safety.
Within the process of designing research and data collection, researchers rely on a sample population — meant to represent the larger population that the research results and other data will be applied to. But in order to be effective, that sample must include all sectors of the larger population. In cases of gender data bias, women are overlooked.
What is gender data bias?
In many areas of research, women are not included in the sample population at an equitable percentage, if at all. This might be fine if the data wasn’t being applied to women, but it is. Products, services and strategies are being generalized to women when the research behind them is not based on data involving women. Read that again: There are things being done and being used by women and we don’t know if they work for or are safe for them. Consider these examples:
Women’s hands are typically smaller than men’s (by about one to two inches), but this is usually not considered when cell phones are designed. These now-virtually mandatory tools — and specifically a trend towards their increasing size — do not take into account how a phone will fit in smaller hands. While some companies offer smaller models, they are typically less powerful or offer fewer options. Another example involves Google Home: The speech recognition databases used to develop this application — according to a 2016 study by sociolinguist and data scientist, Dr. Rachael Tatman — were dominated by male voices, making it 70% more likely to recognize and effectively respond to men’s voices over women’s.
Women are more likely to die of heart attacks because their symptoms are often considered “atypical.” This is because standard symptoms were identified based on research that focused on men’s presentations (chest pain, left arm pain) versus females’ (breathlessness, nausea, fatigue, stomach pain). Collectively, this bias is often referred to as The Yentyl Syndrome, and is detailed in a 2011 article in the European Heart Journal. This makes the male body the default for medical understanding, and the same is true for medical research, in which approximately 85% of rodents used in testing are male, according to a 2011 Neuroscience & Biobehavioral Reviews article.
Women are more likely to be seriously injured in car accidents. Why? Because automobile manufacturers have a “standard seating position” used for safety research that’s based on men’s dimensions. Women are typically shorter than men, and so need to sit higher and closer to the steering wheel to see clearly, but this information is not included in the makers’ “standard.” Women also have a higher likelihood of dying in a car crash because of a similar gender data bias. Male crash-test dummies are typically used for the driver’s seat test. When female crash-test dummies are used, they are commonly limited to the passenger seat. The result is that existing research is neither accurate for nor applicable to female drivers.
In some countries, female officers are wearing vests designed for and researched on male bodies, them more vulnerable and less protected than leaving their male peers. A 2017 Trades Union Congress report detailed these disparities in British PPE application, and many of the same problems apply equally in the United States.
Related: Labelled Women and Bias at Workplace
Research published by Nature Climate Change in 2015 made clear how workspace affects productivity, and likewise demonstrates a gender data bias. As a woman, if you wonder why you are always cold at work while your male counterparts are comfortable, it’s because physiological differences aren’t considered when researching the ideal temperature for employee productivity and comfort. Research tends to use male physiology as the standard, which doesn’t account for sex differences with respect to body mass index or overall body structure. And from a clothing perspective, men tend to wear more suits and layers as part of the typical office attire, while women don’t. When not considered as part of research on office climate, women are left at a disadvantage.
This area of study and its application are typically dominated by men, and are too often limited to their perspectives: data regarding women is largely not considered. As an example, subways are built for efficiency and affordability, and dim lighting and unpopulated areas are more prevalent away from monitored areas. This leaves women at a significant disadvantage, creating spaces where attacks and/or harassment are more likely. In addition, suburbs are still designed using an outdated perspective of the man as the breadwinner and emphasizing daily commute efficiency (according to a 2021 Our Secure Future article). This makes the management of home life (including errands and child care) more challenging, and that vital task often still falls within the responsibilities of women.
How, you might be asking, can there be this rampant disregard for data in so many areas of life? The answer is simple and is the Catch-22 of this situation: Women are considered second, which means they are not even thought of when research is being organized, which then contributes more to women being considered second. It’s also expensive to create the more comprehensive (or multiple) research studies needed to capture all possible consumers connected to the data. This cyclic pattern maintains the status of women as invisible or irrelevant, and keeps them subsumed within male constructs that have dominated, controlled or usurped areas of life and functioning.
What can we do?
Addressing these issues starts by demanding better — regardless of our roles or gender. We can influence the decisions impacting this data bias through actions and choices.
As members of society, we can become critical analyzers of information, including asking about and/or researching how data is collected. Pose questions about who is included in a sample population to learn whether the individuals the data is being applied to were equitably represented. If they weren’t, question the research and challenge businesses by not investing your money in their products.
As leaders, carefully consider your teams — particularly ones creating research, collecting data or utilizing data. Are they diverse, not just in their technical and professional skills, but in who they are? Do they represent the population the data will be applied to? Do they require more diverse and representative perspectives? As a leader, you can also create professional cultures that encourage individuals to challenge protocols and data. Having harmony in group and team discussions is not always ideal, as it doesn’t allow for new perspectives and doesn’t encourage the sharing of concerns.
As women, we can also use our money to influence change. Women are estimated to engage in 70% to 80% of all consumer purchasing, yet many companies continue to use data skewed towards men — to market to men or emphasize their preferences. Women can refuse to put their money into companies that do not embrace equality in research protocols; otherwise, they’ll have little motivation to change.
Ignoring gender data bias contributes and facilitates the placement of women as secondary in our world. Making progress as a society and as women starts with acknowledging how representative research and data can level the playing fields. When we make that happen, we impact women in the present and in the future, and throughout all realms of life.