Project Scoring for Program Evaluation and Teacher Professional Development

Evaluating student learning in K-12 computer science programs is challenging due to the limited number of validated assessments, especially below the high school level. Because computer science is often taught as an engineering-oriented discipline in K-12, it seems natural to use students' programming projects as evidence of student learning. However, this approach is time-consuming and can overestimate what students understand. In this paper, we share our experience with using students' programming projects as a tool for evaluating a middle school computer science program at San Francisco Unified School District. The scoring and analysis of programming projects did generate useful insights about the nature of student learning in the program. More importantly, we found that the process of co-developing a scoring rubric and scoring projects alongside CS classroom teachers generated rich discussions and knowledge-sharing about the current state of the district's CS program, its values, and its long-term goals.