Content area
Experimentation is widely used in fields like marketing, technology, and education, valued for its ability to compare the effectiveness of interventions through randomized, data-driven experiments. Despite its potential, its application in computing education remains underexplored. This thesis investigates how experimentation can be applied more effectively in computing education research, addressing its current usage, challenges, and potential to support evidence-based instructional practices.
This thesis examines the state of experimentation in computing education, focusing on its usage, terminology, and reporting practices. The relevance of these considerations lies in the context of computing education, a field characterized by rapid change in technologies, resources, and student needs. Experimentation offers a scalable, responsive framework to keep pace with these evolving demands while systematically generating actionable insights. To better understand the areas where interventions could be impactful, surveys and interviews explore the evolving challenges faced by computing students, identifying opportunities for targeted, subpopulation-specific interventions. Case studies demonstrate the adaptability of experimentation, including a flipped classroom intervention comparing self-explanation mediums (voice versus text) and email reminders testing variations in tone and content. These studies illustrate how experimentation enables precise comparisons and provides data-driven insights that refine instructional design.
This thesis demonstrates that experimentation is a robust tool not only for measuring instructional effectiveness but also for refining pedagogical practices and fostering a culture of continuous improvement. This work provides a foundation for broader adoption of experimentation in computing education research, fostering practices that are not just evidence-based but also context-sensitive, adaptive, and well-suited to the dynamic needs of diverse learning environments.