Students are organized into teams of five students. During the Fall 2015 semester, we experimented with teams of three. However, teams of this size had two major problems. First many teams of three did not have sufficient collective brain power to solve the problems of the level of difficulty for which we were striving. Teams of five, however, are sufficiently large that they have sufficient brain power to solve more challenging problems. The second issue with teams of three is that in a team of three the team is critically hindered by a member who is absent. In teams of five meanwhile, the absence of one team member still leaves the team with sufficient brain power to solve the problems and complete the labs successfully.
The teams are formed shortly after the semester begins and remain constant throughout the semester. Constant teams provides sufficient time for the teams to gel and really begin to function as a coherent team. Again following Michaelsen, we do not allow the students to form their own teams. Teams are constructed using the iPeer software and the results of this survey. Teams were constructed to be heterogeneous in almost every variable. However, in order to limit the potentially negative effects of soloing, teams were constructed to ensure that women were not alone in their teams – a result more easily achieved in an algebra-based class dominated by life-science majors than in a course for say engineering or physics majors.