Team selection methods for student programming projects

This paper focuses on team selection methods and related issues that may have a substantial impact on students' overall experience working on a “large-scale” programming project. Teams can be selected randomly, by the students, by academic performance, by using psychological profiles, or by using a combination of these methods. Advantages and disadvantages are presented for each method. Issues concerning size of teams include number of students in the class, type and size of software to be built, single project per team versus shared project, software tools available, and the programming skills and general academic maturity of the students in the class. From the students perspective, the most discernable effects of team size involve communication, scheduling, and general cooperation. Important grading issues are described with emphasis on peer grading. The paper concludes with a summary of observations and experiences the authors have had including pitfalls to be avoided.