Effectiveness of Peer-mediated Learning for English-language Learners: A Meta-analysis

Abstract

This dissertation presents the results of a meta-analysis of the effectiveness of peer-mediated learning for English Language Learners(ELLs). The sample included 44 independent studies reporting five distinct outcome types, and analyses included main effects analysis, heterogeneity analysis, analysis of publication bias, and moderator analysis. Results indicate that peer-mediated learning is effective for ELLs across every outcome type analyzed, with estimates for most outcomes proving robust to potential distortions from cluster-randomization, outliers, and publication bias. Few moderators proved significant at the α=.05 level; nonetheless, study quality and post hoc adjustments for pre-test differences were significant for a number of outcome types indicating that study quality variables proved the largest moderators of effectiveness estimates. However, qualitative analyses indicate that effect sizes varied across a number of theoretically important moderators, including: segregated settings, interventions using students’ native languages, students’ age, language learning setting, and language and cultural backgrounds of the students.