The impact of captioning and playback speed on listening comprehension of multilingual English learners at varying proficiency levels

Despite the widespread use of internet-based platforms by multilingual English learners (MELs), research on the intersection of listening comprehension and the ability to manipulate technological features is under-studied. This study explores how the use of captions affects MELs' listening comprehension of easier versus harder questions at three different playback speed levels (0.75, 1, 1.25). It also addresses how overall proficiency, listening subscores (TOEFL), and four background and experiential factors moderate possible effects of captions. The quasi-experimental design involves three cross-sectional experimental and control groups (captions vs. no captions) organized by playback speed levels: Level 1 = slower, Level 2 = default, Level 3 = faster. MANOVA and MANCOVA tests were performed to examine data from 287 MELs who viewed a YouTube TED talk and answered 8 comprehension questions (4 easier, 4 harder). Findings revealed that use of captions had several significant effects on MELs' responses to both easier and harder questions across all levels, with impact being larger at Level 1 than Level 2 and Level 3. They also suggested that overall English proficiency, listening subscores, and several background and experiential factors moderated some main effects of the experiment on participants’ responses. Implications for additional language education, both for the acquisition and instructed contexts, are discussed.