Designing with and for Youth: A Participatory Design Research Approach for Critical Machine Learning Education

As big data algorithm usage becomes more ubiquitous, it will become critical for all young people, particularly those from historically marginalized populations, to have a deep understanding of data science that empowers them to enact change in their local communities and globally. In this study, we explore the concept of critical machine learning: integrating machine learning knowledge content with social, ethical, and political effects of algorithms. We modified an intergenerational participatory design approach known as cooperative inquiry to co-design a critical machine learning educational program with and for youth ages 9-13 in two after-school centers in the southern United States. Analyzing data from cognitive interviews, observations, and learner artifacts, we describe the roles of children and researchers as meta-design partners. Our findings suggest that cooperative inquiry and meta-design are suitable frameworks for designing critical machine learning educational environments that reflect children's interests and values. This approach may increase youth engagement around the social, ethical, and political implications of large-scale machine learning algorithm deployment.