Feminist Data Visualization

D’Ignazio and Klein’s Feminist Data Visualization is not a guide to chart aesthetics so much as an argument about how charts make knowledge. Drawing on feminist science and technology studies, HCI, digital humanities, and critical cartography, the authors reject the idea that visualization simply makes already-complete data transparent. Researchers decide what to collect, which differences become fields, which values count as outliers, and which visual form, interaction, and audience to design for. Those choices shape what can appear as a visible fact and what is excluded before an image is made. Clarity therefore does not automatically mean fairness, and an appearance of objectivity cannot erase a designer’s position or the institutions that shape the work.

The paper’s most usable contribution is a set of six connected principles, each translated into questions about design process and output. Rethinking binaries means asking how categories such as gender are constrained at collection, and why people who do not fit a schema so often become “edge cases.” Embracing pluralism asks designers to disclose their own position and make room for situated accounts that need not collapse into one consensus. Examining power and aspiring to empowerment moves affected communities from the roles of data source or end user toward participation in defining questions, interpreting material, and shaping its presentation. Yet more choice or open access is not automatically empowering: who holds decision-making power, and who bears the risks, still matter. Considering context warns that “cleaning” and standardizing data can also strip away local meaning, histories, and the conditions under which the data were produced.

The remaining principles challenge the assumption that visual evidence is purely rational. Embodiment and affect can be ways of knowing, not merely noise to remove; this does not make feeling a substitute for quantitative evidence, but asks how different forms of knowledge can inform a design. Making labor visible follows data beyond the finished graphic to collection, annotation, cleaning, maintenance, and design: who does this work, who receives credit and resources, and whose contribution disappears? The principles thus concern accountability across problem-setting, collaboration, data handling, representation, and circulation—not a style label applied at the end.

Read the paper alongside the authors’ later Data Feminism and Gender Shades: the former develops a broader account of data and power, while the latter offers a concrete case for examining classification and performance differences. Keep the paper’s own scope in view: the authors present the six principles as a preliminary invitation to further dialogue and practice, not a universally applicable checklist or an empirically validated guarantee of equitable outcomes. Its value is to make design decisions discussable and accountable; responsible judgment still depends on the particular history, data provenance, and knowledge and participation of affected communities.