The Effect of Group Composition on Individual Student Performance in an Introductory Economics Course

Abstract

What’s the best way to allocate students to teams (groups) in those economics courses that rely on small group work to enhance individual student learning? Those who advocate the use of collaborative learning in the college classroom generally suggest combining stronger students with weaker students in three or four member student teams. Many advocates also suggest that there are advantages in forming heterogeneous teams in terms of class year (or age), gender, and ethnicity, as all this helps to provide teams with members who complement each other in skills. Possible drawbacks of such heterogeneity have also been noted. However, little or no empirical work - especially related to the individual performance of students in economics courses - has been done on how best to select the most effective student teams (groups) to enhance individual student learning (as opposed to improving the performance of the group as a whole). This paper hopes to begin to fill this gap. It combines individual observations on 260 students over a three year period in my eight classes of Principles of Economics I at Occidental College, a course that relied quite heavily on collaborative learning in teams of three or four students each. Information from the students’ admissions files (SATS, AP credit, etc.), on the composition of the students’ randomly assigned semi-permanent team, and on their performance on the (identical) final exam, among other information, was used to test the validity of the above advice. Perhaps surprisingly, virtually no empirical support was found, and in some cases the empirical results were directly counter to the above advice. There were other surprising results as well, among them those relating to intercollegiate student athletes (NCAA Division III), and to those students who were the only nonwhite (or female) students on an otherwise all-white (or all-male) team.