Students Struggle to Explain Their Own Program Code

内容摘要

We asked students to explain the structure and execution of their small programs after they had submitted them to a programming exercise. These questions about learner's code (QLCs) were delivered at three occasions in an online and open course in introductory programming as a part of the digital learning material. We make inductive content analysis to research the open-ended text answers we collected. One third of the students struggled to explain their own program code. This estimates possible occurrences of fragile learning at the moment when a student seemingly succeeds in a program writing exercise. Furthermore, we examine correlations between the correctness of the answers with other learning data. Our results indicate that answering properly aligned QLCs correctly has stronger correlation with student success and retention than merely submitting a correct program. Additionally, we present observations on learning event-driven programming to explore QLCs' potential in identifying students' thinking process.