Gender data bias in product and policy design

In our world today, almost everything is designed with men as the default. From the buildings we live in to the policies that govern us, the male perspective is too often the only one considered. This pervasive bias extends even to critical areas like medicine and safety, where the consequences can be life-threatening for women. Take, for instance, the development of drugs. Even when creating treatments specifically for women's health issues, researchers frequently test these drugs on male subjects—both animals and humans. This approach fails to account for the physiological differences between men and women, leading to medications that may be less effective, or even harmful, for women. The repercussions of this oversight are not just about discomfort or inconvenience; they can have serious, even deadly, consequences. But it doesn't stop there. Consider the car you drive. Most crash test dummies used in safety testing are modeled after the "average" male body. As a result, women are 47% more likely to suffer serious injury in a car crash than men. The seats, seatbelts, and airbags are designed to protect a male body, leaving women at a disproportionate risk of harm. This is a systemic issue that permeates nearly every aspect of our lives. The world we live in has been engineered with one gender in mind, leaving the other to navigate a landscape that often doesn't account for their needs. It's not just about fairness, it's about safety, health, and ultimately, survival. It's time for a radical shift in how we design our world. We must move beyond a one-size-fits-all approach and start considering the diverse needs of all people. This means demanding that medical research includes both men and women, that safety standards account for different body types, and that every aspect of our world, from public spaces to the products we use, reflects the reality that women are not an afterthought. Women's lives, health, and well-being should never be compromised simply because the systems in place fail to consider them. It's time to build a world that is truly inclusive, where everyone’s safety and health are prioritized, not just those of half the population. Creator: @annaakana Shot by @johnleestills Grip @meliseeta Sound @mobleywillwork Edited by @benchinapen La Gouvernance au Féminin - Women in Governance

Did you know that in the UK, women are 50% more likely to be misdiagnosed after a heart attack? This is not because women’s symptoms are “atypical”, they are typical for women, but because medical research and training have historically centred on male bodies. The “Reference Man” has been the standard for everything from drug dosages to car crash test dummies, with devastating results: women are 17% more likely to die in a car crash, and 47% more likely to be seriously injured, simply because safety systems were not designed with them in mind. Caroline Criado Perez outlines this concept of 'male as default' in her book, Invisible Women: Data Bias in a World Designed for Men. Even today, the vast majority of pain studies are conducted on male mice, and drugs are often not tested on women at different stages of their menstrual cycle. The result? Women experience more adverse drug reactions, and sometimes, the drugs simply don’t work for them. This isn’t just a healthcare issue, it’s a design issue. When we build systems, products, or workplaces around a single “default” user, we inevitably exclude others. The cost is not just inconvenience, but real harm. In the case of organisations outside of healthcare, these same default user error easily occurs. And this is causing real harm to individuals, teams, companies and has real knock-on effects in society. So what can we do? We must design workplace systems, policies, and cultures that recognise and accommodate difference. We must examine inherent bias that exists in our people processes - right from designing a role, through to the language used in job descriptions and the expectations around what a job 'should' look like. There are so many aspects to eliminating bias in the workplace and this is just one of the starting points for organisations. At Shape Talent Ltd, we've developed a Debias Audit, designed to identify barriers and gender biases in organisations, many of which can be subtle and unintentional. People processes, policies and systems may have been passed down for years without anyone looking critically at them and asking whether they are gender inclusive in their ability to attract, retain, promote and reward talent. Organisations may not be aware of the simple, subtle and often quick changes they can make to their HR policies and practices that positively influence gender equality in the workforce. If this might be a helpful tool, get in touch to discuss how we can customise it for your organisation. #GenderEquity #EDI #DebiasAudit #DEI #PeopleAndCulture #HRLeadership #HRToolkit

The dictionary defines a "gender gap" as when there are disparities and great differences between genders, in many spheres of life: social, economic, cultural, political. But did you know there is also a gender gap in data?   📱 If you are a woman and have ever thought the same smartphone that men seem to easily use with one hand seems unwieldy and simply big in your hand…   💊 If you know of a woman who came into a hospital emergency room complaining of chest pains and was not immediately treated for a possible heart attack...   🚘 If you, as a woman, have struggled to fit comfortably in a car seat belt without adjusting it continuously or needed a cushion on the seat to help you see comfortably over the steering wheel…   …then you get an idea of what the results of a gender data gap look like.   The gender data gap is when data is collected, analyzed, and used as a basis for decision-making without taking into account gender and the associated differences: physical, mental, social, economic, and so on.   That smartphone seems unwieldy because it was designed keeping in mind men's bigger hands. Historically, clinical trials have included fewer women or never separated out the data for men and women to find differences in how men and women are diagnosed or react to medicines. Cars are designed primarily using crash test dummies that fit male proportions and anatomy.    The gender data gap leads to data bias, skewing results and our interpretations of them. And biased data can lead to wrong decisions that can have a negative impact, magnifying disparities. And in today's world where a data driven algorithm seems to rule pretty much everything, that is bad news.   What can you do about it?   ✅ Become aware of any unconscious gender biases when generating or consuming data and its interpretations. For example, if you say human and a man comes to mind, that's an unconscious bias. ✅ If you are a researcher generating data, you can ensure the data you are collecting is representative, i.e. the similar fractions of male and female responses as in the population demographic. And if it is not, add a note indicating it. ✅ If you are a consumer of data, look to see if there is a good representation of males and females. This will help you interpret the results and any decisions based on it with the right perspective. ✅ If you are an entrepreneur or business owner, make sure you are designing your products and services around the needs of both men and women. ✅ If you are a leader, ensure you listen to both the men and women in your teams and take decisions based on balanced views and advice. Do you have more ways to beat the gender data gap? #DataStory #GenderGap #Dareation

When we say “apply a gender lens to everything,” this is exactly what we mean. For over a century, many everyday inventions were designed by men, for men, and women simply adapted, often at a cost to their comfort, safety, and health. The redesigned speculum is a perfect example. After more than 150 years, women engineers finally asked a simple but radical question: What if women’s bodies and experiences actually mattered in medical design? And the speculum is not alone. Inventions originally not designed with women in mind: • Car seat belts – modeled on the “average male body,” increasing injury risk for women and pregnant people • Car crash test dummies – long based almost entirely on male bodies • Smartphones – larger sizes that strain smaller hands • Voice recognition systems – struggle more with women’s voices • Office temperatures – calibrated to male metabolic rates • Personal protective equipment (PPE) – often ill-fitting for women • Medications & drug dosages – tested primarily on men • Tools and machinery – designed for male grip strength and height And here is a critical fact many people still don’t know: 👉 Women were largely excluded from clinical research until 1993. Yet women metabolize drugs differently, experience different side effects, and respond differently to treatments. Still, medicine was labeled “neutral.” It wasn’t. And yes, this goes beyond medicine. Even road and transport design carries gender bias. Without applying a gender lens: • Roads prioritize private cars over public transport • Routes ignore caregiving travel patterns (school, market, home, hospital) • Pedestrian safety, lighting, and crossings, used more by women are undervalued In Ghana, this means some road designs systematically disadvantage women, especially those who walk, carry loads, use public transport, or move with children. I’ll write more about this when I get the time, but this is the heart of the matter: Design is never neutral. Policy is never neutral. Research is never neutral. If gender is ignored, inequality is designed in. #academia #phdjourney #genderlens #womeninacademia #scicom

⭐ Today we unveil key insights from our upcoming Outlook Study on #AI and #Gender. Our findings reveal that current AI #policyframeworks often overlook gender considerations. Notably, the Global Index on Responsible AI (GIRAI) indicates that gender equality is one of the lowest-scoring areas in government frameworks. Out of 138 countries assessed, only 24 mention gender in AI, and a mere 18 address it significantly. The study also highlights the specific #risks women face from biased AI systems, such as recruitment #algorithms favoring male candidates. According to MIT’s AI Risk Repository, issues affecting women are predominantly categorized under Discrimination and Toxicity, emphasizing biases that lead to stereotyping and marginalization. Moreover, the lack of gender-disaggregated data hampers our ability to assess the effectiveness of interventions. Our study combines in-depth analysis, real-world examples, and actionable policy recommendations to expose how biases disproportionately affect women, revealing systemic barriers to gender equality in AI. Through the implementation of the Recommendation on the #Ethics of AI and the W4EAI network, UNESCO is committed to driving meaningful change. To foster a more inclusive AI landscape, we must confront these challenges head-on and advocate for diversity and equitable outcomes for all. Let's work together to create a better future!

🔍 If you don’t see the problem, you’re probably not the one affected by it. The world we live in wasn’t designed with women in mind—it was designed around men. Not intentionally, but because data, research, and policies have historically ignored half the population. Caroline Criado Perez’s Invisible Women highlights just how deeply embedded these biases are: 💔 Heart attacks look different for women. But since most medical research is based on men, women are 50% more likely to be misdiagnosed. ❄️ Office temperatures? Set for men. Women often freeze in workplaces because climate settings are based on the metabolic rate of a 70kg man. 📱 Your phone? Too big. Tech is designed to fit the average male hand, making it harder for women to use comfortably. 🚗 Crash-test dummies? Male by default. Women are 47% more likely to be seriously injured in car accidents because safety testing doesn’t account for their bodies. ⏳ The unpaid labour gap? Enormous. Women globally do 75% of unpaid care work—limiting their career opportunities and economic independence. These aren’t minor inconveniences. They are life-threatening, career-limiting, and entirely avoidable—if we start designing a world that includes everyone. Now, with AI, these biases aren’t just persisting—they’re being automated and amplified at scale. International Women’s Day isn’t just about celebrating achievements. It’s about challenging the systems that still don’t work for half the population—and ensuring AI doesn’t make it worse. 📢 Time to close the gender data gap—before AI locks it in.

Who Gets to Design the World We All Live or Die In? The THOR-05F NHTSA milestone means: it's Story Time over here. Early in my tenure at Tesla, I was pulled into a late-stage pre-launch test for our prototype SUV, the Model X. I came in on the weekend, not because it was part of my job, but because the vehicle dynamics team realized: 💡 “We haven’t tested this iteration of the car with a small woman passenger yet.” I became the stand-in for half the population at the end of the process, when design decisions were already locked. That moment has stayed with me. The data shows why this matters: ⚫ Women 9% more likely to die in car crashes (down from ~17% in 2013) ⚫ 73% more likely to be seriously injured ⚫ Crash tests used male dummies as the default for decades ⚫ Only 25% of the automotive manufacturing workforce is women (even less in engineering workforce) This isn’t about intent. It’s about who’s in the room. I share this because last month, the National Highway Traffic Safety Administration NHTSA hit another progress milestone on the Long Road to using representative female crash test dummy. It's a story that rears it's head with another not-yet-complete-solution every ~8 months. And it’s not just cars 🚗. Phones, medical devices, safety gear, and norotiously: office temperatures are all on the long list of “Design Defaults” for all-gender products designed around only-male users. This isn't a women’s issue. It's a design failure that directly causes real-world safety threats. Two options to de-risk: 1.) The process fix says: "Literally just invite some end-users to sit in the room in some early development conversations! It's not hard. Invite an expansive set of them! More 'flies on the wall' to influence assumptions up front make a huge difference -- rather than at the end, when you just use them to validate decisions already made." And it's true that talking to customers & users early is important. BUT, you'll still depend on a perfect process to not "miss" some use cases. 2.) The "team DNA" fix is more elegant: "Include a variety of end users inside your dev team." Literally hire engineers who represent the world you're creating for. (for example, the coolest Mom Car on Earth needs some engineering leads who are Moms. (Shoutout to the 2 that were!)) ➡️ Thinking expansively requires systems thinking, research, and creativity – but the less homogenous the team (in professional experience & in personal user-like behavior), the more likely you can get away with diligence and process misses, without facing serious failure. ➡️Better yet, you get more expansive thinking at *every* step of development. Expansive teams create more creative, more correct products. That’s how we design a world that works right.

What Happens When Women Lead in Data? We talk a lot about data, how to collect it, analyze it, and leverage it. But rarely do we ask: Who’s leading these efforts? When women lead in data science and AI, they bring more than technical skills. They bring perspective, empathy, and a drive to build systems that work for everyone. And the results are Powerful, Inclusive and Transformative. Here’s how; 1. Inclusive Design Take Femtech as an example. Alicia Chong Rodriguez, founder of Bloomer Tech, developed a smart bra that monitors women’s heart health, something sorely missing from mainstream medical tech. Why:- - Because most health data excludes women. - Women in leadership noticed this gap, and innovated around it. When women lead, blind spots get solved. 2. Tackling Bias from the Inside Dr. Joy Buolamwini founded the Algorithmic Justice League to challenge racial and gender bias in AI. Her work exposed how facial recognition systems perform worst on women with darker skin. Because the training data was biased. The system followed suit. Her leadership pushed Big Tech to reform.   Lesson: Data doesn’t lie, but it often reflects our existing biases. 3. Building Trust with Ethical Leadership Women leaders tend to drive more transparent, people-centered decision-making. This isn’t about being “soft.” It’s about building trust with users, teams, and the public. In the age of AI, where transparency is everything, this is a leadership advantage. 4. Better Representation When women lead data initiatives: Data models become more inclusive. Assumptions are challenged. Outcomes become more equitable. It’s not just a win for women, it’s a win for innovation and society at large. These wins aren’t unicorn stories. They’re evidence of what’s possible when we make space for women in data leadership. Let’s stop treating them as exceptions, and start seeing them as the standard we should all aspire to. 👉 Who’s a female data leader that inspires you? Tag her and let her know she’s making a difference.