Data feminism begins from a precise refusal: data cannot be treated as neutral evidence detached from the social world that produces it. The chapter opens through Christine Darden’s work at NASA, where mathematical expertise, racial hierarchy, gendered labour and national technological ambition converge in one scene. The iconic idea is situated data: every dataset is made somewhere, by someone, under institutional conditions that decide what can be counted, who is credited, and whose labour disappears behind the authority of calculation. The chapter therefore shifts data science from technical procedure to epistemic politics. Data does not simply represent reality; it participates in organizing reality through categories, absences, classifications and visual forms. A feminist approach does not weaken objectivity by adding identity or politics. It strengthens knowledge by forcing it to account for power. This means asking who benefits from a model, who is harmed by a classification, whose histories are erased by aggregation, and whose expertise remains invisible because it does not fit the dominant image of technical authority. The text is essential because it transforms feminism into a method for better knowledge: rigorous because accountable, empirical because situated, critical because it treats data as a social relation rather than a purified instrument.
