Data, Design and Democracy : Embedding the Ethics of Algorithms into Data Science Curriculum

• In Event: Research Infrastructures, Digital Tools and New Directions in STS Research

Abstract

This paper shares accounts of ways data ethics and the politics of algorithms can be embedded in a transdisciplinary data science program to transform data practice. Critical data studies brings into the open a core message that data does not speak for itself, but rather, is given a voice by the people, algorithms and structures that play increasingly critical roles in the transformation of data into insight. Using a social informatics and STS perspective to shape curriculum sensitises students to the significance of the co-evolving relationship between people and data technologies for their work practices. To embed in our students an awareness of the consequences of choosing specific technologies for their practice, we examine the infrastructures within which those data practices unfold. Students engage at a theoretical level with questions of how and where data systems and classifications are produced alongside the practicalities of the data work they perform in real-life, project-based contexts. Their critical reflection on these socio-technical entanglements sheds light on the relations of power and invisibility and the implications of varying perceptions of visibility for their practice. The paper shares the conceptual framework, learning design that contributes to students’ deep engagement with the background “shadow work” of their knowledge practices, and accounts from students’ blogging about implications for their practice. Early experience with this approach suggests that such theoretical engagement has the potential to shape data science practice. By foregrounding critical data ethics considerations, students are inspired to become agents of change in their data practices.

Author

• Theresa Dirndorfer Anderson, University of Technology Sydney