Professional Development for High School Computer Science Teachers on Data Science through a Feminist Lens

This Research-to-Practice Work in Progress paper presents a recent professional development for high school computer science teachers. The court system increasingly relies on algorithms to rate a criminal defendant's likelihood of recidivism. These algorithms utilize machine learning models tuned on biased datasets, resulting in false assessments. There are several other alerting examples of how a software engineer's personal bias affects their design and development of technology. This bias leads to poor usability of technology for underserved groups. Data science applies machine learning on massive datasets to invent new technology to improve everyday life from healthcare to consumer needs. Designing and developing data science solutions for a diverse population requires an unbiased approach. Data feminism, which emerges from the intersecting fields of data science and feminist theory, presents a methodology to solve data science problems given three principles: 1) invent new ways to represent data unknowns, 2) invent new ways to reference the material economy behind the data, 3) make dissent possible. In this paper, we present a case study on a week-long online professional development for computer science high school teachers on data science using a data feminist lens. We discuss the data science tools and technology taught during the weeklong event. We share examples of the data feminism principles and class exercises. The computer science teachers also participated in community building activities. The culminating weeklong challenge was to analyze an earthquake dataset, which included themes of socioeconomic disparity within a city. The teachers worked in teams to explore this dataset with the new data science techniques and recommend resource management solutions for a city's earthquake response. In this paper, we explore the question: how the participating computer science high school teachers incorporate data feminism into their data science pipeline? We analyze the audio transcripts collected during the week to create visual analytics on our findings. We consider the impacts of culturally responsive approaches to computer science K-12 curriculum and future professional development in this space.