Effectiveness of peer-mediated learning for English language learners: A meta-analysis

Abstract

This manuscript reports the findings from a series of inter-related meta-analyses of the effectiveness of peer-mediated learning for English language learners (ELLs). Peer-mediated learning is a broad term that as operationalized in this study includes cooperative learning, collaborative learning, and peer tutoring. Drawing from research on interaction in second language acquisition, as well as from work informed by Vygotskian perspectives on socially-mediated second language learning, these meta-analyses synthesize the results of experimental and quasi-experimental studies. Included studies were conducted with language learners between the ages of 3 and 18 in order to facilitate comparisons to US students in K-12 educational settings. All participants were identified as ELLs, though learners in both English as a Second Language (ESL) and English as a Foreign Language (EFL) settings were included. Similarly, learners from a variety of language backgrounds were included in order to facilitate generalizations to the linguistic diversity present in US schools, and abroad. Main effects analyses indicate that peer-mediated learning is effective at improving a number of outcome types, including: language outcomes, academic outcomes, and social outcomes. Funnel plots and Egger’s regression analyses were conducted to examine the probability of publication bias, and it appears unlikely in most analyses. Moderator analyses were also conducted, where sample sizes were sufficient, to examine what measured variables were capable of explaining heterogeneity in effect sizes between studies.