Preservice elementary teachers who were assigned teacher roles prior to a unit on probability showed greater achievement than those assigned student roles. When students did not know to which roles they would be assigned prior to class there was no difference between achievement of "teachers" and "students." (SD)