The Effectiveness of Multilingual Versus Monolingual Approaches to L2 Teaching and Learning: A Meta-Analysis of Learning Outcomes

Abstract

The debate on multilingual (codeswitching) versus monolingual (immersive, target-language only) approaches to second language (L2) teaching has received renewed attention in recent decades. Both approaches are employed in L2 teaching and learning, supported by competing philosophies regarding the need for maximal L2 input and the facilitative potential of first language (L1) inclusion. Given a growth in experimental research comparing the effectiveness in terms of learning outcomes of the two approaches in L2 classrooms, this meta-analysis study investigates the effects of multilingual versus monolingual approaches on outcomes as measured by language assessments and examines the variables (age, proficiency, learning context, target skill area, and procedure for learning) that moderate such effects. We conducted a comprehensive search for empirical investigations of the topic, resulting in 24 studies comprising 30 samples with a total of 2,138 learners after the application of inclusion and exclusion criteria. Through aggregating the effect sizes from this sample, analyses revealed that compared with a monolingual approach, a multilingual approach better facilitated L2 learning with a weighted between-group effect size of d = .49 and a within-group effect size of d = 1.63. The analyses of moderator variables revealed that the positive effects on the learning of skills and language content areas of a multilingual approach appear to hold across all ages and proficiencies, especially for beginning-level learners and in conditions where both teachers and students employed their L1 in L2 teaching and learning. Also, a multilingual approach showed a larger effect in studies where speaking was the target skill area, or where the treatment was longer, was conducted in foreign language settings, was conducted in intact classrooms, or was administered by a teacher or teacher-researcher. Our findings contribute evidence supporting the effectiveness in terms of learning outcomes of a multilingual (codeswitching, translanguaging) over a monolingual (immersive, L2-only) approach and provide insights into the nuances involved. These include the skill areas and proficiency levels benefiting the most, and the importance of involving teachers as agents of research in their classrooms.