“Invisible women: Data bias in a world designed for men” by Caroline Criado Perez

This award-winning book explains how the failure to include women in data sets and decision-making has major health and safety effects and leads to missed opportunities for potentially transformative insights.

 By Patricia A. Maurice with Janet G. Hering

12 November 2024, DOI: 10.5281/zenodo.14024426

We women in STEMM love data, numbers, and statistics. Caroline Criado Perez’s “Invisible women: Data bias in a world designed by men” [1] is chock full of concrete, quantitative examples of why our long history of overlooking women matters. Males – their bodies, their ideas, their experiences, and their needs – are considered as normal and universal whereas women are considered as unusual or ‘niche.’ Females are literally half the world’s population, but we are often considered anomalous, too difficult to study, and too hard to accommodate.   Although this book is about five years old, it is still worth reading not just for ideas and examples but because it points out sources of data that are collected on a recurring basis and cites many useful papers. These can serve as entry points for more advanced searches for literature and/or data.

Lack of data including women or of gender-disaggregated data has profound effects on women’s health and well-being. For example, women’s heart attacks can go undiagnosed because their symptoms are different from men’s.  Medications can be useless or even harmful for women because they were tested only on (mostly young) men.  Women often have different transportation needs than men, related to different care-giving roles, jobs, accessibility to family-shared cars, and vulnerability to violence, including sexual assault. Failure to consider women’s bodies, activities, and needs in designing automobiles, buses, roads, and parks can lead to dangerous situations and inefficiencies.  The extent of women’s exclusion from data creates a cycle of discrimination; this ‘data gap’ needs to be fixed.

The book also articulates how ‘When we exclude half of humanity from the production of knowledge we lose out on potentially transformative insights.’ A problem – whether social, political, scientific, or mathematical – that seems intractable and unsolvable to men can often be solved with the help of female perspectives.  Women’s participation in politics has been found to correlate with greater investment in education in several countries; the converse has also been observed.  In this regard, it is particularly problematic that women’s political ambitions are often perceived as unseemly, as Criado Perez highlighted in the case of Hillary Clinton.      

Criado Perez directly addresses challenges women face in the workplace, including in academia, and provides statistics on inequities in ‘unpaid work’ performed by women versus men.  She gives examples of how addressing issues for women in the workplace also helps men, as we also discussed in our post on ‘It’s not pie!’ [2].

Patricia’s personal reflections on the book

My first thought about the book was: If everyone read it, perhaps more people would understand why using inclusionary words actually matters and how everyone would benefit from a more inclusive world. 

My second thought was that, if I were a guy with a female partner, mother, sister, or daughter, I would want to fill the data gaps and create a more equitable world. A typical man or boy, anywhere in the world, is likely relying on at least one woman to do a lot of unpaid work. Without saying that this is a good thing, women’s health and productivity obviously benefits men. When women are employed in a paid job, it’s good for the entire family if she receives a decent and equitable wage.  I have not heard these arguments articulated very often, but men rely on women so much that it makes sense for men to care about the damage done by data gaps.

As a professor emerita in STEMM, the book highlighted how important it is to teach STEMM students about the effects of data gaps and to include women in design of experiments, data collection, and technical specifications.  Senior women leaders can help to ensure that the data and design gaps get filled and that medicine and technology work for women and girls, as well as for men.  We are educators and researchers, and we have a major role to play in solving these problems.

Reading the many stories about women being ‘invisible’ made me think more about how my own life has been impacted over the years.  As a girl, I was often told about all the things I couldn’t be good at because I had a female brain.  It was confusing because I wasn’t supposed to be good at math or science, but I was.  I delayed pursuing a PhD because I never saw examples of women succeeding in math and science, especially in academia.  If a movie like ‘Hidden Figures’ had come out when I was a kid, it might have made a big difference to my self-confidence and ambition. Throughout history, there have been women who were successful in art, literature, science, math, etc. but their accomplishments have long been ‘invisible.’ 

I remember being assigned to read Elaine Morgan’s “The Descent of Woman” [3] by a female instructor in a human paleontology course years ago. I’m appalled to find that when I googled ‘human evolution’ today, the first ~40 illustrations are totally male.  In fact, I had to search long to find one illustration that included evolution to modern women. So, while books like “Invisible Women” are having an impact on tech industry portrayals (google ‘engineer’ or ‘scientist’), the problems persist.

One of the most powerful concepts in the book is that the female body is hyper-visible to men when it is sexualized but invisible in instances where visibility might actually help women.

The book reinforced my belief that senior women leaders can make a real difference, for everyone.  There is a discussion of problems with work environments having been designed for men with stay-at-home wives, including problems with the academic tenure system. Criado Perez’s discussion reminded me of all the times when having just one senior woman leader in the room, especially with a male ally, made a huge difference in dealing with an unhelpful policy. Women were not the only beneficiaries.

Finally, I am more concerned than ever about how a world obsessed with ‘big data,’ artificial intelligence, and machine learning is going to be impacted by gender-related data gaps and the ‘invisibility’ of women. Many of the software engineers and data scientists developing the algorithms are young men. Most of the data available for training machine learning projects contains gender- and or race-related data gaps as shown in the film “Coded Bias” [4]. Recently, I observed gender bias when using AI-driven drawing software.  When I wrote instructions to sketch a physics professor lecturing a class or a scientist working in the laboratory, the AI-generated illustrations were white men (see graphic above [5]). To get broader representation, I needed to specifically use the words ‘woman’ or ‘female’ or ‘African American.’ Criado Perez’s book was written before ChatGPT burst upon the world. We need more studies of how historical and ongoing gender bias may be affecting AI-derived products.

Conclusions and questions for further thought

My main conclusions are that this book is well worth reading and that I hope Caroline Criado Perez will write a new edition soon.

Some questions for further thought:

·       In your own research and teaching, do you specifically address potential gender bias?

·       Have you read the book or watched any interviews with the author?  If so, how were your perceptions changed?

·       In what ways do you think you, personally, are ‘visible’ or ‘invisible’?

References and notes 

[1] “Invisible Women: Data bias in a world designed for men” by Caroline Criado Perez, 2019, United States, Abrams.

[2] https://www.epistimi.org/blog/its-not-pie-how-equity-for-women-in-stemm-can-benefit-everyonenbsp-nbsp

[3] “The Descent of Woman” by Elaine Morgan, 1972, Souvenir Press, UK

[4] “Coded Bias” is a 2020 film by director Shalini Kantayya (https://www.codedbias.com/).

[5] The figure at the top of this post was AI generated in June 2024 using Stable Diffusion v1.5 on the Draw Things app and the simple description ‘physics professor”

Previous
Previous

What we can learn from “The Guerrilla Girls: The Art of Behaving Badly”

Next
Next

Preparing to lead effectively through emerging challenges