"Data Feminism" offers a thought-provoking and transformative perspective on how we collect, analyze, and interpret data. Authored by Catherine D'Ignazio and Lauren F. Klein, the book presents a new framework for thinking about data science and its impact through the lens of feminism. This approach challenges the traditional, often patriarchal narratives that dominate the field, advocating for a more equitable and inclusive data science practice.
The book is structured around seven principles of data feminism, which include examining power, challenging inequality, and embracing pluralism. These principles guide the reader through a series of compelling case studies and examples that illustrate how data can be used to reinforce or challenge societal norms. For instance, the authors examine what are by now issues in biased data that most of us are aware of, such as medical studies that are conducted only on cismen, with serious health consequences for AFAB (assigned female at birth) patients. However, the authors go well beyond the obvious, looking at how we collect and even how we define data is often filled with biases. We define paid labor as part of the economy, for instance, while ignoring the unpaid childcare, eldercare, and support labor of which women do the majority. They also stress the importance of context in analyzing data - for example, if we state that women are underrepresented in STEM careers, for instance, without addressing issues that reduce the pipeline of femme scientists, we are opening up the data to being interpreted, as many on the right have, as indicative that women lack interest or capacity for such work. D'Ignazio and Klein argue that data is not neutral; it is shaped by the conditions under which it is collected and the biases of those who collect and interpret it. Therefore, they call for a more reflexive approach to data science that acknowledges and addresses these biases.
One of the book's strengths is its non-technical nature. Despite dealing with complex issues at the intersection of technology, gender, and society, Data Feminism is written in a clear and engaging style that makes it more a work of social critique than a math book, which it emphatically isn’t. This readability does not come at the expense of depth or rigor; the authors support their arguments with a wealth of research and examples from a variety of disciplines.
Data Feminism also stands out for its actionable recommendations for both data analysts and users of data. The authors not only critique the current state of data science but also offer concrete suggestions for how individuals and organizations can practice data feminism. These suggestions range from how to conduct more inclusive data analyses to how to create data visualizations that reflect a diversity of experiences. One drawback of the book is a definition of women that includes the standard trope that seeing the world through an empirical lens is masculine and regarding it with an emotional eye is feminine. Another frame, that the mind is masculine and the body, feminine, is also presented, albeit with nuance. The chapter encouraging “emotional, embodied” interpretations of data as being more in line with “female” ways of thinking sat uneasily with me, as it may to many contemporary feminists tiring of stereotypes, even when they are attired in empowerment. However, this chapter also makes vital points, such as the importance of understanding that all reality is filtered through individual perspectives, and that no data-driven approach to a topic conveys omniscience on its users.
Even with these limitations, Data Feminism is a must-read for anyone interested in the intersection of data science and social justice. It challenges readers to rethink their assumptions about data and provides a compelling roadmap for creating more just and equitable data practices. Whether you are a data scientist, a feminist scholar, or simply someone interested in the ethical implications of technology, this book offers valuable insights and practical advice for navigating the complex world of data with a critical, feminist viewpoint.