Designing Problem Sessions for Algorithmic Subjects to Boost Student Confidence

Abstract

In this paper, we describe how we changed the structure of problem sessions in an algorithmic subject, in order to improve student confidence. The subject in question is taught to very large cohorts of (around 900) students, though our approach can be applied more broadly. We reflect on our experiences over a number of years, including during the pandemic, and show that by adding clear sectioning indicating the style of the questions and by including simple warm-up questions, student engagement and confidence improves, while making the teaching activities of our teaching assistants easier to manage.