On Students’ Experiences with Algorithm Tracing using Pair Programming

We have seen students struggling with fully understanding the algorithms when required to trace pseudocode for an Algorithmics course, even though examples and exercises on tracing each algorithm are provided throughout. This course is being taught in year 2 of the new Graduate Apprenticeship in Software Engineering programme (GA) [2], covering fundamental string and graph algorithms and basic computability topics.

Code tracing helps students develop valid mental models of a program [7], while sketching [5, 9], used for tracing code via pen-and-paper drawing visualising program states or other computing processes, may help students manage cognitive load while understanding the notional machine [8] behind an algorithm. Hence sketching is fitting for tracing algorithm pseudocode.

Pair programming (PP) [3] has been successfully adopted in CS education as a collaborative learning activity [6] where both learners work in pairs to solve a task, with one being the driver and the other the navigator, and swapping roles regularly. In a previous investigation on using PP for constructing finite state automata [1], we have found that GA students enjoyed working in pairs.

We aim to investigate the feasibility of employing PP for algorithm tracing, similar to its application in programming. Our initial research question is: How do GA students experience the use of pair programming for tracing pseudocode of string and graph algorithms?

Algorithms were taught over two weeks in block-mode, with one-hour lecture followed by a one-hour tutorial session each day, and 2-3 tracing exercises per tutorial. Students were encouraged to use PP for all tracing exercises during tutorials. In March 2023 we ran a pilot study to gather initial feedback on the PP setup. We asked students’ opinions on what they liked and what they didn’t like about using PP for algorithm tracing via an anonymous, online survey consisting of two open-ended questions. Seven out of 25 students enrolled in the course responded. The inductive thematic analysis of the survey responses shows that most students enjoyed collaboration to build up their understanding of algorithms. Themes around insufficiently supportive setup and time-consuming activity have been identified.

The results of this pilot study suggest that while the activity was perceived as valuable, the PP setup requires enhancements. Future actions include: increase tutorial sessions structure with timing guidelines for each tracing exercise; provide blank paper and/or digital templates for tracing each algorithm; regularly assess progress in pairs and encourage students to seek assistance when encountering difficulties; utilise information from previous courses that use PP to pair up students with similar computational thinking skills [4]. To support students who opt to not use PP, we will explore methods to actively engage them. This includes identifying and utilising computer-based tools to assist with visualising algorithm execution and offering additional tutor support.

We will continue using PP for tracing algorithms in the upcoming teaching session and further investigate the students’ experiences and their perceived and assessed performance when using PP.