Objective Data? The Ways That Gender Bias Can Enter the Data Pipeline

Data are used everywhere: ranging from Healthcare to Marketing and Finance, the way we analyze information holds considerable influence in society. In many fields, data analytics has the potential to either sustain prevailing societal inequalities or transform them – however, data are often riddled with biases. To stimulate discussion around this topic and connect stakeholders, Myrthe Blösser and Paulina von Stackelberg started FemData. With FemData, they specifically focus on the ways that gender biases can distort the model outcomes by influencing the collection, labeling, and interpretation of data, leading to skewed insights and perpetuating inequality. In this talk, Myrthe and Paulina introduce their initiative and provide examples from previous research on the topic of gender bias in data applications. And what potential solutions are there to mitigate bias

About the speakers

Myrthe Blösser is a PhD candidate in consumer psychology with the Marketing group of Amsterdam Business School. In her research, she examines different tools and regulations to help citizens with low AI literacy protect themselves against unfair and discriminatory AI.

Paulina von Stackelberg is a PhD candidate in Business Analytics at the University of Amsterdam. Her research revolves around prediction models for time-dependent data in healthcare and renewable energy. In the context of her data science work, she is also interested in the social implications of data applications.

Berty Bannor (moderator) With her background in public international law, Berty works as a staff member at Bureau Clara Wichmann, a feminist organization in the Netherlands. The organization advocates for legal and social change on a variety of issues to promote gender equality and women’s rights through strategic litigation, research, and advocacy. Within the organization, Berty focuses on themes such as work and income, as well as tech and online safety for women.