Investigating the effects of whole-class singing activities on linguistic outcomes of young foreign language learners in English primary schools

Thesis

Cultural beliefs in songs' effectiveness for teaching young foreign language learners are common, despite scant credible empirical evidence to support such intuitions. This two-phase thesis investigates the empirical evidence underpinning teachers' pedagogical choices regarding the use of songs in foreign language teaching for achieving linguistic outcomes. It then proposes a coherent and systematic way forward for research evidence to better support teachers' judgements.

Phase 1 systematically reviews intervention research comparing the effects of using songs to other pedagogical tools on the linguistic outcomes of young learners in formal second or foreign language education. This phase appraises 60 studies conducted from 1978–2021, finding the cumulative evidence weak and inconclusive across all measures.

Phase 2 builds on the findings of the systematic review by implementing an intervention study with 96 beginner French learners (age 7–8) in two primary schools in England to assess the substantive effects of singing on oral language development through an elicited imitation task. It adopted a randomised controlled trial design to compare the effects on this outcome of presenting French through songs, chants or stories with each other and with an active control. Participants received 242 minutes of French input in 11 lessons over three weeks. Their elicited imitation task performance was measured at pretest, posttest and delayed posttest. Data were transcribed, scored from 0 (omission) to 5 (exact imitation), and analysed using a cumulative link mixed model. Performance improved over time for all groups, with significant interaction effects between song and control groups at posttest, and between all treatment groups and control at delayed posttest. No significant interaction effects were detected between treatment groups. Bayes factors were also calculated, finding insufficient evidence to support the null hypothesis of no difference between conditions.

This thesis empirically interrogates longstanding positive bias toward believing songs constitute effective SLA pedagogy. It contributes empirical evidence on the effects of songs on young learners' linguistic outcomes and enhances our understanding of using songs with beginner language learners in England's primary schools, both under-researched areas. The findings have implications for teaching and future research on using songs to teach young language learners.

Authors: Hamilton, C

• DOI: 10.5287/ora-5zj9650vk
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: MFL; elicited imitation; oral language development; young learners; songs; applied linguistics; foreign language learning; systematic review; RCT; chants; randomised controlled trial; primary school languages; music; nursery rhymes
• Subjects: Education, Primary; Children's songs, French; Applied linguistics; Language and education