Implementing EarSketch: Connecting Classroom Implementation to Student Outcomes

The expansion of computer science into more classrooms invites researchers and evaluators to shift their focus from predominantly measuring student-level factors to measuring both student- and classroom-level variables. Research presented in this article uses multi-level modeling to study student-level factors within the larger context of classroom-level factors. Specifically, we analyze EarSketch, a collaborative and authentic learning tool, that introduces students to programming through music remixing, has previously been shown to increase student engagement, and increases learners' intentions to persist in computing. This article presents classroom implementation frameworks commonly used in math and science education but rarely, if ever, applied to computer science. The results from a multi-level modeling analysis show that classroom implementation correlates with students' intentions to persist in computing but may not be related to student attitudes toward computing or content knowledge acquisition. Further analysis reveals that one of the five classroom implementation factors, elaboration, emerges as the most salient. This article triangulates these results with qualitative findings from school administrators and teachers, and the article concludes by theorizing how classroom implementation frameworks may be adapted to meet the unique needs of computer science teachers, learners, researchers, evaluators, and curriculum developers.