Abstract:

Active learning has become the leading evidence-based instructional method over traditional lecture-based methods, improving students’ long-term retention and problem-solving abilities. Large and diverse introductory Computer Science (CS) classes pose a unique challenge for implementing active learning, as students in the same class can be years apart in programming experience. Thus, standard active learning activities such as multiple-choice questions and code-writing exercises are frequently too easy for some and too hard for others, providing valuable learning gains only to those in a narrow band of skill levels. Moreover, they fail to teach students the skills needed for real-world software engineering, such as reading, debugging, and giving feedback on other people’s code.

This work presents a toolbox of 10 different activities that engage learners across the spectrum of skill levels. Some activities challenge advanced students with optional activities and competitive extensions, and others support struggling learners by scaffolding the process of starting to code. The activities presented also teach students essential expert competencies like code style, efficiency, and systematic testing, and give them ample opportunities to analyze and learn from their peers’ code.

The activities were piloted in MIT’s introductory level CS class, a large lecture-hall class with an enrollment of around 350 students, and evaluated by a survey completed by 60. The exercises were divided in two categories: technical exercises, where students answered a question or solved a coding-related task, and metacognition exercises, where students reflected on their learning. According to the survey, 89% of students found the technical exercises helpful to their learning, and 58% found the metacognition exercises helpful. Participation data reveals that these active learning exercises engage learners equally across all levels of prior coding experience. The findings demonstrate that an innovative approach to active learning fosters deep engagement, even in large and diverse classrooms.

Keywords:

Computer Science, Education, Active Learning.