Investigating student-instructor interactions when using pair programming: An empirical study

Abstract

At North Dakota State University, there are multiple sections of the CS1 and CS2 introductory computer science courses. A large number of students are enrolled in each section, making it difficult to hold laboratory sessions as there is not enough space for all of the students in one room. This results in diminished student attendance and a decrease in student understanding of the course material. Pair programming has been shown to have multiple benefits in educational use. Previous research has shown that it benefits student learning in addition to increasing the student retention in computer science programming courses. Using pair programming would also allow students to share laboratory resources and make it possible to accommodate more students in laboratory session. To study the effects of pair programming on student-instructor interactions in laboratory sessions of introductory computer science courses, an empirical study was conducted at North Dakota State University. Data about student-instructor interactions was collected by monitoring the laboratory sessions during the study run, as well as through a post-study survey given to students and interviews with the instructors. The results from this study indicate that having students work in pairs as opposed to individually reduces the number of questions from students and decreases the amount of time that a student must wait for instructor assistance.