Association for Computing Machinery
ITiCSE 2018 Proceedings of the 23rd Annual ACM Conference on Innovation and Technology in Computer Science Education
Assessing the impact of regional or statewide interventions in primary and secondary school (K-12) computer science (CS) education is difficult for a variety of reasons. Qualitative survey data provide only a limited view of impacts, but quantitative data can be notoriously difficult to acquire at scale from large numbers of classrooms, schools, or local educational authorities. In this paper, we use several publicly available data sources to glean insights into public high school CS enrollments across an entire U.S. state. Course enrollments with NCES course codes and local descriptors, school-level demographic data, and school geographic attendance boundaries can be combined to highlight where CS offerings persist and thrive, how CS enrollments change over time, and the ultimate quantitative impact of a statewide intervention. We propose a more appropriate level of data aggregation for these types of quantitative studies than has been undertaken in previous work while demonstrating the importance of a contextual aggregation process. The results of our disparate impact analysis for the first time quantify the impact of a statewide Exploring Computer Science (ECS) program rollout on economic groups across the region. Our blueprint for this analysis can serve as a template to guide and assess large-scale K-12 CS interventions wherever detailed project evaluation methods cannot scale to encompass the entire study area, especially in cases where attribute heterogeneity is a significant issue.