Stakeholders' Interpretations of Data for Equitable Computing Education


Computing education has growing inclusion and equity challenges (e.g. exclusionary online learning experiences, biased assessments, inadequate student feedback mechanisms). Many groups experience minoritization in computing education, including students who are Black, Indigenous, and people of color (BIPOC), women, non-binary students, transfer students, international students, first-generation students, and students with disabilities. To ensure diverse students can realize their potential to participate in and challenge computing communities, we must enable stakeholders (e.g. students, teachers, curriculum designers) to take informed, timely, and equitable actions. This dissertation explores how to design interactions with data to inform stakeholders in support of such actions.

While data often perpetuates and exacerbates inclusion and equity challenges when improperly used, it can also support equity-oriented goals if properly contextualized for interpretation by stakeholders. I explored how stakeholders interpreted data in three contexts: 1) informing students of what to learn next in an adaptive, self-directed online learning experience; 2) informing curriculum designers with empirical evidence of assessment bias; 3) and informing teaching teams of inequities using contextualized student feedback. Through these studies, I identified how stakeholders’ relationships with educational structures and systems impacted their interpretations of data for equity-oriented goals. These factors have implications to the research and practice of learning at scale, computing education, and human-computer interaction. Therefore, I claim the following thesis statement:

Interactions with data that consider prior knowledge, perceptions of power relationships, and cultural competency can enable computing education stakeholders to connect their interpretations of data with their domain expertise in service of equity-oriented goals.