Experimental Analyses of Peer Tutoring: Toward a Technology of Generative Learning

Numerous studies since the 1960s have demonstrated that peer tutoring between two students is an effective teaching practice across populations and academic content areas. However, there has been limited research on peer tutoring beyond the traditional dyad format. We analyze variations of peer tutoring in a series of studies in a 3rd grade general education classroom. During the first study, we compared dyad and group peer tutoring structures for spelling acquisition with 14 participants across 2 experiments. We found that the majority of the students mastered novel words with fewer learn units during peer tutoring in dyads (i.e., lower learn units-to-criterion). However, scores on spelling post-assessments were higher following group peer tutoring conditions. Findings suggest that peer tutoring in a group may be more effective when considering post-assessment accuracy. In the second study we analyzed the effects of peer tutoring with Equivalence Based Instruction (EBI) on inference making, specifically, the emergence of eight 3-member fraction-percentage classes with 8 participants across 2 experiments. We found that participants acquired both baseline training relations during peer tutoring EBI: one directly as a tutee and one indirectly as a tutor. Following peer tutoring EBI, all participants derived the remaining 4 relations. Once participants had formed equivalence classes, they could also sort fraction stimuli, which demonstrated transfer of function. Additionally, we noted that it may be important for instructors to consider response effort for training relations when designing instruction for peer tutoring EBI due to possible adverse effects on student behaviors. Our findings suggest novel and effective means in designing pedagogy to increase learn units, select effective tutoring formats, and plan for inference making in general education classrooms