Outcome feedback during collaborative learning: Contingencies between feedback and dyad composition

The role of outcome feedback in collaborative learning settings has received little empirical attention. We examined whether outcome feedback improves learning gains in singleton and dyadic learning conditions, while specifying different dyadic pairing options. In a randomized experiment, 496 ninth-graders solved challenging tasks that required fully developed proportional reasoning to be solved correctly. Based on individual pretest performance, each student was assigned to one of three levels of proportional reasoning competence (Wrong₁, Wrong₂ and Right) and randomly assigned to either work alone or with a (Wrong₁, Wrong₂ and Right) peer. Half of the dyads and singletons were given the opportunity to empirically test their solutions and received outcome feedback from an objective testing device. The results indicated that when collaboration is considered as a general condition, learners in dyads and singletons profited equally from outcome feedback. When different dyadic compositions are specified, however, the combination of collaborating with a "Right" partner and receiving outcome feedback proved to be particularly powerful. Outcome feedback did not improve learning in any of the other conditions. Furthermore, and contrary to the "two-wrongs-make-a-right-effect", interaction between two different "Wrong" students did not yield larger gains than other pairing options. The outcomes are discussed in light of existing theories and research.