Presentation

How much training is enough? A time-course analysis of a 10-session audiovisual HVPT for L2 English vowels

---

Abstract

High-variability phonetic training (HVPT) effectively improves second language vowel perception, yet little is known about how learning unfolds across training sessions or how much training is optimally required. This study examines the time course of perceptual learning in a 10-session audiovisual HVPT program targeting British English monophthongs. The 10-session design was selected to ensure sufficient exposure to the full vowel inventory, in line with extended HVPT protocols in previous research. Thirty-two advanced Spanish-Catalan bilingual learners of English completed 10 training sessions over six weeks, with vowel identification accuracy tracked session by session. Learning trajectories were captured through peak accuracy and learning rate (early and late learning slopes). Mixed-effects models and cluster analyses were used to examine overall learning trajectories, learner profiles, and vowel subset differences. Results showed significant gains in perceptual accuracy, with the largest improvements occurring early and peak performance reached around Session 7. Thereafter, learning plateaued, with only limited or unstable gains in Sessions 8–10. Cluster analyses revealed four learner profiles—high achievers, early high performers, late improvers, and strugglers—differing primarily in learning rate and peak timing rather than final attainment. Front vowels were consistently identified more accurately than low and back vowels, which remained more resistant to training across profiles. These findings indicate that HVPT gains are nonlinear and learner-dependent, suggesting that adaptive training designs adjusting duration to learner progress may optimize outcomes.

Keywords: speech perception, HVPT, audiovisual speech perception, pronunciation instruction, high variability training, second language, vowels

How to Cite:

Aliaga-García, C. (2026). How much training is enough? A time-course analysis of a 10-session audiovisual HVPT for L2 English vowels. In P. Trofimovich, A. Bodea, T.-N. N. Le, R. Suzuki, & C. Teló (Eds.), Proceedings of the 16th Pronunciation in Second Language Learning and Teaching Conference (pp. 1–17). Concordia University, Montreal, Canada, October 2025. https://doi.org/10.31274/psllt.22955