摘要:What is the best way to allocate students to small teams in those economics courses that rely on small group work to enhance individual student learning? While experts in collaborative learning provide many suggestions, little empirical work has been done. This article begins to fill the gap. It examines whether a variety of characteristics of the small semipermanent team to which an introductory economics student was randomly assigned affect the student's performance on the identical final exam in eight sections of a principles course over three years, holding individual student characteristics the same. The empirical results suggest that peer effects from such group characteristics have weak or negligible effects and do not support the experts' general advice on how to form teams. A few other perhaps surprising results emerged.