Partnership Characteristics and Student Performance in an Introductory Computer Science Course

Group work and programming in partnerships have been shown to have a positive effect on student learning in computer science education. As a result, students in introductory computer science courses often work in teams or partnerships. In this paper, we examine the composition of student partnerships on programming projects in an introductory computer science course and its relationship to student performance. We analyze data from 1,434 students enrolled in an introductory computer science course at a highly ranked public university, collected over one academic year. These data include student group composition, performance on projects and exams, and demographic information. The course is a second-semester “CS2” programming and introductory data structures course, and is part of the core computer science curriculum for majors and minors. Students had the option to work in partnerships or to work alone on projects. Our results examine several factors influencing the success of a partnership, including difference in cumulative grade point average (GPA), gender balance, and work habits like starting projects early. After controlling for GPA, we observed an association between starting projects early and increased performance on both exams and projects. The impact was greatest among those in the lowest GPA quartile, where an early start made the difference between an average final letter grade of C+ (lowest early-start quartile) and B(highest early-start quartile).