Evaluating the Efficacy and Impacts of Remote Pair Programming for Introductory Computer Science Students

With the increase in online learning, instructors are looking for novel ways of supporting student learning and getting students to collaborate in online environments. Pair programming allows students to brainstorm and problem-solve in teams and has been found to help with improving code design, attitudes toward computer science, productivity, and performance. However, past work has focused on face-to-face, in-person collaboration, and it is unclear whether these benefits will translate to an online context. This work replicates several studies evaluating the effects and benefits of in-person pair programming in an online environment. In an introduction to programming course, students participated in weekly online sessions where they were asked to solve a set of exercises in pairs or individually. We measure task performance and student opinions on the activities and perceptions of remote pair programming. Our study found that remote pair programming had little to no impact on the time taken, promising but not statistically significant impacts on code correctness, and statistically significant impacts on students' perceptions of both their own experience and the efficiency and efficacy of pair programming. Our findings show that some, but not all, of the benefits of pair programming can be replicated in an online context.