2024 ASEE Annual Conference & Exposition

Self-reflection can improve one’s understanding and processing of complex concepts. It stimulates deep-learning and professional development. In this work, we introduced self-reflection assignments as an instrument to aid students, in the machine learning course under investigation, to prepare for the course conceptual-oriented exams since the programming homework assignments mainly focus on the practical aspects and applications of the studied algorithms. ANCOVA analysis was conducted to statistically study the effect of self-reflection assignments on exam scores. The analysis was conducted on two exam scores from two different cohorts, one of which did not use the self-reflection assignments while the other used reflection assignments to review the course contents before the exams. The results suggest that there is insufficient evidence to substantiate the claim that student performance in exams has shown improvement as a direct consequence of utilizing self-reflection assignments. In addition, an anonymous student survey was used to study how the students perceived the self-reflection assignments and how it helped deepen their understanding of the course materials and prepare them for taking the exam. The majority of the students who completed the survey had a positive perception of the reflection assignments. They found that the assignments helped solidify their high-level understanding of the different algorithms taught in class and aided in recalling the materials as they prepared for the exams. In future iterations of the course, we plan to revise the language and frequency of the assignments to more effectively assess students' understanding of the theoretical concepts.